NAMIC Wiki - User contributions [en]

DBP2:Harvard

2010-02-11T19:21:55Z

Dougt: /* Other Harvard-NAMIC Collaboration Projects */

Back to [[DBP2:Main|NA-MIC DBP 2]]
__NOTOC__
= Overview of Harvard DBP 2 =
== Velocardiofacial Syndrome (VCFS) as a Genetic Model for Schizophrenia ==

VCFS is a genetic disorder characterized by a deletion of a small piece of chromosome-22. The features of this syndrome include deficits in neurological psychomotor and perceptual skills, as well as in cognitive domains such as learning and memory. Most importantly, up to 30% of VCFS patients develop schizophrenia, making it the most commonly known single risk factor for the development of psychosis and a unique model for studying neurodevelopmental changes leading to psychotic deficits. We plan to collect new, high resolution DTI, structural and fMRI data, and apply existing NAMIC tools, as well as help to develop new tools to investigate the contribution of genetic variation to brain and behavioral/cognitive abnormalities, thus bridging the gap between neuroimaging studies and genetics.
The NAMIC community will gain access to de-identified imaging data (new, high resolution structural and diffusion data acquired on the 3T magnet at Brigham and Women’s Hospital). Unlike in schizophrenia, subjects with VCFS have concrete cognitive abnormalities, in addition to a well defined chromosomal abnormality, which, taken together, will make it easier to establish scientific protocols that reveal associated anatomical and functional brain abnormalities in this disorder. Interestingly, some anatomical abnormalities will be shared between VCFS and schizophrenia (e.g., connections within working memory circuits), and some will be not (e.g., sensory and motor paths). Neuropsychological and genetic data will also be collected for each individual as part of separate collaboration between PI and the Children's Hospital("Investigation of Genotype/Phenotype Correlations in Velocardiofacial and DiGeorge Syndromes”) and de-dentified dataset containing neuropsychological tests, clinical evaluations and genetic data will be provided to the PI. The NAMIC project with thus enable the PI to apply NAMIC tools to imaging VCFS data, a genetic disorder that is viewed as a genetically mediated subtype of schizophrenia. To date, there have only been a small number of neuroimaging studies of this disorder and no studies have combined neurocognitive, neuroimaging, and genetic investigations in the same study. Importantly, this research will also increase our understanding of schizophrenia, and will help establish a multimodal research project involving an important collaboration between computer scientists, cognitive neuroscientists, radiologists, psychiatrists, and geneticists. The focus on imaging and genes also affords a new window of opportunity for defining further the new area of “imaging genomics”. [[DBP2:Harvard:Introduction|More...]]

Data is provided at the following link: '''[[Data:DBP2:Harvard|Harvard Data]]'''.

= Harvard Roadmap Project =

{| cellpadding="10" border="1" style="background:lightblue;text-align:left;"

| style="width:200px" | [[Image:Arcuate_new.png|200px]]
| style="width:1300px" |

== [[DBP2:Harvard:Brain_Segmentation_Roadmap|Stochastic Tractography for VCSF]] ==

The main goal of this project is to develop end-to-end application that would be used to characterize anatomical connectivity abnormalities in the
brain of patients with velocardiofacial syndrome (VCFS), and to link this information with deficits in schizophrenia. [[DBP2:Harvard:Brain_Segmentation_Roadmap|More...]]

<font color="red">'''New: '''</font> 12/10/2008 official release of the python Slicer 3 stochastic tractography module

<font color="red">'''New: '''</font> Collaborative Publication: "Reduced interhemispheric connectivity in schizophrenia-tractography based segmentation of the corpus callosum. Kubicki M, Styner M, Bouix S, Gerig G, Markant D, Smith K, Kikinis R, McCarley RW, Shenton ME. Schizophr Res. 2008 Dec;106(2-3):125-131. Epub 2008 Sep 30.

| style="width:200px" | [[Image:Corpus2.jpg|200px]]

|}

= Other Harvard-NAMIC Collaboration Projects =
; <font color="firebrick" font size="4"> GA Tech </font>

{| cellpadding="10" border="1" style="background:lightgray;text-align:left;"

| style="width:1700px" |

== [[Projects:OptimalMassTransportRegistration|EPI distortion correction using optimal mass transport.]] ==
EPI distortion correction using optimal mass transport. (baseline vs. strct t2w). Goal is to use optimal registration for EPI distortion correction in diffusion weigthed scans.

|-
| |

== [[Projects:GeodesicTractographySegmentation| Geodesic Tractography Segmentation]] ==
We are currently investigating Cingulum Bundle fractional anisotropy (FA) differences between a population of 12 schizophrenics and 12 normal controls using this new methodology.

[[http://pnl.bwh.harvard.edu/pub/papers_html/MelmohanMICCAI07.html Finsler Tractography for White Matter Connectivity Analysis of the Cingulum Bundle J. Melonakos, V. Mohan, M. Niethammer, K. Smith, M. Kubicki, A. Tannenbaum miccai 2007 ]]

|-
| |

== [[Projects:TubularSurfaceSegmentationPopStudy| Tubular Surface Segmentation Population Study]] ==
We are currently investigating Cingulum Bundle white matter properties between a population of schizophrenics and controls using using the Tubular Surface Model.

[[http://www.na-mic.org/publications/item/view/1571 Niethammer M., Zach C., Melonakos J., Tannenbaum A. Near-tubular fiber bundle segmentation for diffusion weighted imaging: segmentation through frame reorientation. Neuroimage. 2009 Mar;45(1 Suppl):S123-32. PMID: 19101640. PMCID: PMC2774769. ]]

|-
| |

== [[Projects:LabelSpace| A Coupled Multi-Shape Representation]] ==
We are currently using this technique to build an unbiased atlas for segmentation.

[[http://users.ece.gatech.edu/~malcolm/pubs/malcolm_lss.pdf J. Malcolm, Y. Rathi, and A. Tannenbaum. "Label Space: A Multi-Object Shape Representation." In Combinatorial Image Analysis, 2008.]]

|-
| |

== [[Projects:MultiscaleShapeAnalysis| Shape Analysis of the Caudate]] ==
Multiscale shape analysis

[Y. Gao, D. Nain, M. Styner, M. Niethammer, J. J. Levitt, M E Shenton, G Gerig, A. Bobick, A. Tannenbaum. Framework for the Statistical Shape Analysis of Brain Structures using Spherical Wavelets. In preparation for the Insight Journal, February/March 2007]
|}

; <font color="firebrick" font size="4"> MIT </font>

{| cellpadding="10" border="1" style="background:lightgray;text-align:left;"

| style="width:1700px" |

== [[Projects:ShapeBasedSegmentationAndRegistration|Shape Based Segmentation and Registration]] ==
We use this segmentation algorithm to process our entire data set.

[[http://pnl.bwh.harvard.edu/pub/papers_html/PohlIEEE07.html A Hierarchical Algorithm for MR Brain Image Parcellation, K. Pohl, S. Bouix, M. Nakamura, T. Rohlfing, R. McCarley, R. Kikinis, W. Grimson, M. E. Shenton, W. Wells, IEEE Transactions on Medical Imaging, Volume 26, Number 9, Pages 1201-1212, 2007 ]]

|-
| |

== [[Projects:GroupwiseRegistration| Groupwise Registration]] ==
The goal is to create unbiased atlases throught nonlinear (b-spline) group wise registration.

[[http://www.spl.harvard.edu/publications/item/view/1090 S.K. Balci, P. Golland, M.E. Shenton, W.M. Wells III. Free-Form B-spline Deformation Model for Groupwise Registration. In Proceedings of MICCAI 2007 Statistical Registration Workshop: Pair-wise and Group-wise Alignment and Atlas Formation, 23-30, 2007. ]]

|-
| |

== [[Projects:DTIClustering| DTI Fiber Clustering and Fiber Based Analysis ]] ==
The goal of this project is to provide structural description of the white matter architecture as a partition into coherent fiber bundles and clusters, and to use these bundles for quantitative measurement. We use this method in two ongoing projects- corpus callosum segmentation in chronic schizophrenia sample, and whole brain clustering in sample of first episode psychosis subjects.

[[http://pnl.bwh.harvard.edu/pub/papers_html/odonnellAJNR06.html A Method for Clustering White Matter Fiber Tracts; L. O'Donnell, Kubicki M, M. E. Shenton, M. Dreusicke, W. E. L. Grimson, C.-F. Westin, AJNR Volume 27, Number 5 2006 ]]

|-
| |

== [[Projects:DTIModeling| Fiber Tract Modeling, Clustering, and Quantitative Analysis ]] ==
The goal of this project is to segment and parametrize fiber bundles for more precise group comparison. We use this method to segment and compare between controls and chronic schizophrenia population tracts belonging to inferior semantic processing stream (Inferior Longitudinal fasciculus, Uncinate Fasciculus and Inferior Occipito-Frontal Fasciculus). Manuscript is in preparation.

<font color="red">'''New: '''</font>[[http://pnl.bwh.harvard.edu/pub/papers_html/madahmiccai08.html Findings in Schizophrenia by Tract-Oriented DT-MRI Analysis, M. Maddah, M. Kubicki, W. Wells, C.F. Westin, M.E. Shenton, W.E. Grimson, miccai, Pages 917-924 September, 2008 ]]

|-
| |

== [[Projects:fMRIClustering| fMRI clustering]] ==
The goal of this project is to use resting state fMRI data and clustering algorythm to detect and separate specific functional brain networks. We are in the process of runing the method on group of chronic schizophrenia and matched control subjects.

|}

; <font color="firebrick" font size="4"> UNC </font>

{| cellpadding="10" border="1" style="background:lightgray;text-align:left;"

| style="width:1700px" |

== [[Projects:ShapeAnalysisFrameworkUsingSPHARMPDM| Shape Analysis Framework using SPHARM-PDM]] ==
["Shape Abnormalities of Caudate Nucleus in Schizotypal Personality Disorder" has been accepted by Schizophrenia Research. James J. Levitt, Martin Styner, Marc Niethammer, Sylvain Bouix, Min-Seong Koo, Martina M. Voglmaier, Chandlee C. Dickey, Margaret A. Niznikiewicz, Ron Kikinis, Robert, W. McCarley, Martha E. Shenton.]
<font color="lightgray"> </font>
|}

; <font color="firebrick" font size="4"> Utah 1 </font>

{| cellpadding="10" border="1" style="background:lightgray;text-align:left;"

| style="width:1700px" |

== [[Projects:DTIProcessingTools| Diffusion Tensor Image Processing Tools]] ==
Testing and use of eddy current correction for our DWI scans. <font color="lightgray"> </font>

|}

; <font color="firebrick" font size="4"> Utah 2 </font>

{| cellpadding="10" border="1" style="background:lightgray;text-align:left;"

| style="width:1700px" |

== [[Projects:DTIPopulationAnalysis| Population Analysis from Deformable Registration]] ==
This project uses non-rigid registration of DTI images to produce a common coordinate system for hypothesis testing of diffusion properties. <font color="lightgray"> </font>

|}

;
== Registration Documentation ==

{| cellpadding="10" border="1" style="background:lightgray;text-align:left;"

| style="width:1700px" |

== [[Projects:DBP2:Harvard:Registration_Documentation| Documentation of Slicer3 registration modules]] ==
This page documents the results of using the various registration methods available in Slicer 3. <font color="lightgray"> </font>

|}

PNL SoftwareEngineer

2009-07-27T20:57:03Z

Dougt:

The National Alliance for Medical Imaging Computing (NA-MIC) is a multi-institutional, interdisciplinary team of computer scientists, software engineers, and medical investigators who develop computational tools for the analysis and visualization of medical image data. The purpose of the center is to provide the infrastructure and environment for the development of computational algorithms and open source technologies, and then oversee the training and dissemination of these tools to the medical research community.

We are seeking applications for a software engineering position. The successful applicant will be working alongside a team of neuroscientists investigating the impact of schizophrenia and Velocardiofacial syndrome on brain morphology and function.

Job description:

Image Computing Applications Engineer:

*Summary: The engineer will work closely with local user community to integrate NA-MIC Kit software into local research workflows. The engineer will work with the NA-MIC algorithm and software developers to develop implementation strategies to meet the needs of local users by adapting and/or extending existing code in the NA-MIC Kit. Examples of the engineering work required include writing code to integrate several filters into a pipeline, building user interfaces to algorithms, converting data file formats, and automating repetitive analysis tasks.

*Duties:
**Define and specify user requirements for new software.
**mplement solutions in collaboration with other NA-MIC personnel.
**Support use of new solutions by local users.
**Disseminate solutions to broader community through adherence to NA-MIC software practices including publicly readable repositories, design material and documentation on public wiki, participation in email lists to support users and developers.

*Qualifications: BS Computer Science or equivalent. Knowledge of medical image processing a plus.

*Other Requirements:
**Willingness and ability to develop "production" quality code adhering to existing design standards.
**Knowledge of C++ essential.
**Cross-platform software development experience (Windows and Linux; Mac a plus).
**Experience with NA-MIC Kit tools a plus (CMake, VTK, ITK, KWWidgets, 3D Slicer).

Please see http://www.na-mic.org/pages/Driving_Biological_Projects for more information on our NAMIC collaborations.

Submit resumes and cover letters to sylvain@bwh.harvard.edu

PNL SoftwareEngineer

2009-07-27T20:53:33Z

Dougt: Created page with 'The National Alliance for Medical Imaging Computing (NA-MIC) is a multi-institutional, interdisciplinary team of computer scientists, software engineers, and medical investigator...'

The National Alliance for Medical Imaging Computing (NA-MIC) is a multi-institutional, interdisciplinary team of computer scientists, software engineers, and medical investigators who develop computational tools for the analysis and visualization of medical image data. The purpose of the center is to provide the infrastructure and environment for the development of computational algorithms and open source technologies, and then oversee the training and dissemination of these tools to the medical research community.

We are seeking applications for a software engineering position. The successful applicant will be working alongside a team of neuroscientists investigating the impact of schizophrenia and Velocardiofacial syndrome on brain morphology and function.

Job description:

Image Computing Applications Engineer:

Summary: The engineer will work closely with local user community to integrate NA-MIC Kit software into local research workflows. The engineer will work with the NA-MIC algorithm and software developers to develop implementation strategies to meet the needs of local users by adapting and/or extending existing code in the NA-MIC Kit. Examples of the engineering work required include writing code to integrate several filters into a pipeline, building user interfaces to algorithms, converting data file formats, and automating repetitive analysis tasks.

Duties:
Define and specify user requirements for new software.
Implement solutions in collaboration with other NA-MIC personnel.
Support use of new solutions by local users.
Disseminate solutions to broader community through adherence to NA-MIC software practices including publicly readable repositories, design material and documentation on public wiki, participation in email lists to support users and developers.

Qualifications: BS Computer Science or equivalent. Knowledge of medical image processing a plus.
Other Requirements:
Willingness and ability to develop "production" quality code adhering to existing design standards.
Knowledge of C++ essential.
Cross-platform software development experience (Windows and Linux; Mac a plus).
Experience with NA-MIC Kit tools a plus (CMake, VTK, ITK, KWWidgets, 3D Slicer).

Please see http://www.na-mic.org/pages/Driving_Biological_Projects for more information on our NAMIC collaborations.

Resources

2009-07-27T20:49:56Z

Dougt: /* Active */

=== [[Collaborator:Resources|Resources for Collaborators]] ===

* This page contains information for investigators who would like to collaborate with NAMIC.

=== Data ===

All NA-MIC Data is available at the following link: [[Data|NA-MIC Data]]

=== Software: NA-MIC kit ===

The NA-MIC Kit consists of all software that is being made available under the NA-MIC project. This software follows the NIH guidelines for open software development. In this section, we provide information about the components of the NA-MIC kit as well as supporting software tools that are being used by the software developers on the project.

* [[NA-MIC-Kit|Software Resources for NA-MIC Kit]]
* [[Engineering:SandBox|Development Sandbox ]]

=== Publications Guidelines and Resources ===

The [[Publications:Main|publications page]] contains information on publications guidelines for NAMIC, the funding acknowledement text, as well as the acknowledgements/references associated with each of the data sets.

=== Mailing Lists ===

These are the mailing lists associated with NA-MIC. If you are a participant in the project, please make sure that you are signed up for all the mailing lists that apply to your role and interests in the projects. These lists are moderated and maintained by Kitware.

* [http://public.kitware.com/cgi-bin/mailman/listinfo/namic-algo NAMIC Algo]
* [http://public.kitware.com/cgi-bin/mailman/listinfo/namic-algo-pi NAMIC Algo PIs]
* [http://public.kitware.com/cgi-bin/mailman/listinfo/namic-all NAMIC All]
* [http://public.kitware.com/cgi-bin/mailman/listinfo/namic-bio1 NAMIC Bio1]
* [http://public.kitware.com/cgi-bin/mailman/listinfo/namic-bio2 NAMIC Bio2]
* [http://public.kitware.com/cgi-bin/mailman/listinfo/namic-developers NAMIC Developers]
* [http://public.kitware.com/cgi-bin/mailman/listinfo/namic-dissemination NAMIC Dissemination]
* [http://public.kitware.com/cgi-bin/mailman/listinfo/namic-eng NAMIC-Eng]
* [http://public.kitware.com/cgi-bin/mailman/listinfo/namic-leadership NAMIC Leadership]
* [http://public.kitware.com/cgi-bin/mailman/listinfo/namic-mgt NAMIC Mgt]
* [http://public.kitware.com/cgi-bin/mailman/listinfo/na-mic-project-week NAMIC Project Week]
* [http://public.kitware.com/cgi-bin/mailman/listinfo/namic-service NAMIC-Service]
* [http://public.kitware.com/cgi-bin/mailman/listinfo/namic-sitepis NAMIC Site PIs]
* [http://public.kitware.com/cgi-bin/mailman/listinfo/namic-training NAMIC Training]
* [http://public.kitware.com/cgi-bin/mailman/listinfo/namic-dti NAMIC-DTI Community]
* [http://public.kitware.com/cgi-bin/mailman/listinfo/namic-shapeanalysis NAMIC Shape Analysis Community]

=== [[NIH-Page|NIH Page]] ===

* This page contains useful information provided by our NIH officers.

=== [[Mbirn:Main_Page|Morphometry BIRN Page]] ===

* This page contains information about the [http://www.nbirn.net Morphometry Biomedical Informatics Research Network] Project

=== NA-MIC Powerpoint ===

* [[Media:NA-MIC_Powerpoint_Template_2.ppt|NA-MIC Powerpoint Template]]
* [[Media:NAMIC-Intro-Feb-04-2005.ppt|NA-MIC introduction slides]]

=== [[NAMIC_Logos_Templates|NAMIC Logos and Templates]] ===

* This page contains links to files containing the NA-MIC logo and templates.

=== Job Openings ===
====Active====
* [[2009-IGT-Prostate-Postdoc|Postdoctoral research associate in image processing for Image Guided Prostate Procedures]]
* [http://www.cs.queensu.ca/~gabor/OpenJobs/ITK-Programmer.htm Image-Guided Surgery Applications Engineer at the Perk Lab, Queen's University, Canada]

* [[Summer Intern at GE Research]], post resume at [http://www.ge.com/careers www.ge.com/careers] Job #1001997
* [[PNL_SoftwareEngineer|Software Engineer opening at the Psychiatry Neuroimaging Lab of Brigham & Women's Hospital/Harvard Medical School, Boston, MA]]

====Expired====
* Closed: [[SPL-Postdoc|Post-doctoral Fellow in Radiology, Surgical Planning Laboratory, Brigham and Women's Hospital and Harvard Medical School]]
* [[Slicer-Tester|Medical Image Computing Software Open Source Engineer]]

=== Wikis ===

We are often asked about mediawiki and other wikis. Here is some [[Information_on_wikis|information on wikis]].

2009 Summer Project Week FunctionalClusteringAnalysis

2009-06-26T13:49:48Z

Dougt: /* Key Investigators */

__NOTOC__
<gallery>
Image:PW2009-v3.png|[[2009_Summer_Project_Week#Projects|Projects List]]
Image:Dougt_NAMIC1.jpg|Whole brain clustering in sagittal view.
Image:Dougt_NAMIC2.jpg|Whole brain clustering in axial view.
</gallery>

==Key Investigators==
* BWH: Doug Terry (dougt@bwh.harvard.edu), Marek Kubicki, Andrew Rausch, Lauren O'Donnell, David Tate, CF Westin

<div style="margin: 20px;">
<div style="width: 27%; float: left; padding-right: 3%;">

<h3>Objective</h3>
We are using the clusters that are automatically outputted from whole-brain tractography and whole-brain congealing/clustering to draw correlations based on anatomical connectivity to see if there are group differences between chronic schizophrenics and controls.

</div>

<div style="width: 27%; float: left; padding-right: 3%;">

<h3>Approach, Plan</h3>

After whole brain tractography is completed for each case, the tracts for all the cases will be put through group coregistration/congealing. Then clustering takes place on the mean congealed brain. Specified number of clusters here is 200 for the whole brain, bilaterally. The clusters from the mean image are then transferred back to individual subject space. The mean FA/Mode/Trace will be taken from each cluster on each subject.

These diffusion values will then be correlated with clinical/neuropsych assessments. This is a way to look at possible deficits in schizophrenics without grouping the entire anatomical cluster together.

</div>

<div style="width: 40%; float: left;">

<h3>Progress</h3>
Each case has undergone whole-brain tractography. Group coregistration, congealing, and clustering has taken place and the results have been confirmed as anatomically accurate by Lauren O'Donnell. The most relevant neuropsychological and clinical measures are being selected by David Tate. Correlations between these values and diffusion measures will be run on automatically outputted clusters to see if there are any differences between the 2 subject groups.

For analysis, we are first looking at the data in 2 ways: 1) Within the Normal Controls, we will be measuring FA/Mode/Trace/Axial Diffusion/Radial Diffusion for all the 200 bilateral tracts and correlating them with neuropsychological tests since trends would be expect in a healthy population. 2) Divide each of the 200 clusters down the midsagittal plane to separate the bilateral clusters into left and right sides. Then correlate with neuropsych.

*End of Project Week Update*
Diffusion vales are being extracted from the automatically clustered tract files so that Analysis 1 can be completed in the near future. We're also in the process of writing a script to divide the tract files to separate the tracts in the left hemisphere and the right hemisphere so unilateral analyses can be run.

</div>
</div>

<div style="width: 970%; float: left;">

==References==
* [http://lmi.bwh.harvard.edu/papers/papers/odonnellThesis06.html O'Donnell L. Cerebral White Matter Analysis Using Diffusion Imaging. Ph.D. thesis, Massachusetts Institute of Technology, 2006.]
* [http://pnl.bwh.harvard.edu/pub/papers_html/WhitfordMysell2009.html Whitford T, Kubicki M, King R, Khan U, Markant D, Alvarado J, McCarley R, Shenton ME. Abnormalities in tensor morphology in patients with schizophrenia: A dti study of the corpus callosum, 2009.]

</div>

Events:TutorialContestJune2009

2009-06-25T13:49:02Z

Dougt: /* Tutorials */

__NOTOC__

Back to [[2009_Summer_Project_Week]]

[[Image:Arctic_Logo.png |250px|thumb|right|First Prize in the January 2009 round: [[UNC_ARCTIC_Tutorial|ARTIC Tutorial]] ]]

[[Image:PW2009-v3.png|300px]]
=Background=

[http://www.slicer.org Slicer3] is now being used to perform meaningful research tasks. As part of the NA-MIC Training Core activities we are building a curated portfolio of tutorials for the basic functions and functionality available in Slicer. Examples for such existing tutorials are the level 1 and 2 courses in the [http://wiki.na-mic.org/Wiki/index.php/Slicer3:Training#Training_Compendium|NA-MIC training compendium].

=Tutorial Contest Goal=
The primary purpose of this contest is to enrich the training materials that are available to end-users and developers using 3D Slicer and the NA-MIC kit. We believe contestants will be motivated to participate to enhance the dissemination of their own algorithms that they have incorporated into the Slicer3 platform and/or to enhance training of Slicer3 functionality for their own laboratory groups.

There will be two categories:
#'''END TO END SOLUTION TUTORIAL:''' In this category, the tutorial will teach a user how to solve a particular clinical problem using the NA-MIC Kit. Entries into this category will require at least:
#*materials about the scientific and application background and motivation,
#*step-by-step guides, and
#*sample data
#*Example: [[Media:ARCTIC-Slicer3-Tutorial.pdf|‏ ARTIC (Automatic Regional Cortical Thickness) Tutorial]]
#'''ALGORITHM TUTORIAL:''' In this category the tutorial will teach a user how to make an algorithm work on their data. Entries into this category will require at least:
#*materials about the scientific and application background of the algorithm(s) and their use in the Slicer environment
#*step-by-step guides, and
#*at least two different sample data sets from two different institutions
#*Example: [[media:EMSegment_TrainingTutorial.pdf| Non-human Primates Segmentation Tutorial]]

=Template=
A basic template has been used for all of the tutorials. The same design should be used for the contest. It can be found here: [[Media:TrainingTutorialTemplate.ppt|Template]]
*Note: The examples above predate the template.

=Rules=
*Tutorial must be based on a snapshot or release of Slicer 3
*Tutorial must follow the guidelines specified above
*If applicable, provide clear directions for downloading and installing additional modules
*The tutorial and all of its components (data, powerpoints/pdfs, additional modules etc.) must be released under the [http://www.slicer.org/slicerWiki/index.php/Slicer:license Slicer license]
*Applicants must agree to work with the NA-MIC Training and Dissemination Cores to curate their submission (we will test it on each of the available platforms and for usability and work with you to smooth any issues after the contest).

=Dates=
*Presentation: all tutorials will be presented by the authors on Thursday, June 25th 9-11am during the Project Week. Each tutorial presentation should be 10 minutes long.
*Decisions announced: Friday June 26, 2009, 10 am <br>

=Tutorials=
* [[media:Microscopy-Confocal-TrainingTutorial-2009JUNE.pdf|Confocal Microscopy]]
* [[media:ARCTIC-Slicer3-Tutorial.ppt‎|ARCTIC: Automatic Regional Cortical ThICkness]]
* [[media:DBP2JohnsHopkinsTransRectalProstateBiopsy_TutorialPres2009June.pdf‎|Trans-rectal MR guided prostate biopsy]]
* [[Media:Stochastic_June09_1.ppt|Python Stochastic Tractography Module]]

2009 Summer Project Week FunctionalClusteringAnalysis

2009-06-22T17:57:58Z

Dougt: /* Key Investigators */

__NOTOC__
<gallery>
Image:PW2009-v3.png|[[2009_Summer_Project_Week#Projects|Projects List]]
Image:Dougt_NAMIC1.jpg|Whole brain clustering in sagittal view.
Image:Dougt_NAMIC2.jpg|Whole brain clustering in axial view.
</gallery>

==Key Investigators==
* BWH: Doug Terry (dougt@bwh.harvard.edu), Marek Kubicki, Andrew Rausch, Lauren O'Donnell, David Tate, CF Westin

<div style="margin: 20px;">
<div style="width: 27%; float: left; padding-right: 3%;">

<h3>Objective</h3>
We are using the clusters that are automatically outputted from whole-brain tractography and whole-brain congealing/clustering to draw correlations based on anatomical connectivity to see if there are group differences between chronic schizophrenics and controls.

</div>

<div style="width: 27%; float: left; padding-right: 3%;">

<h3>Approach, Plan</h3>

After whole brain tractography is completed for each case, the tracts for all the cases will be put through group coregistration/congealing. Then clustering takes place on the mean congealed brain. Specified number of clusters here is 200 for the whole brain, bilaterally. The clusters from the mean image are then transferred back to individual subject space. The mean FA/Mode/Trace will be taken from each cluster on each subject.

These diffusion values will then be correlated with clinical/neuropsych assessments. This is a way to look at possible deficits in schizophrenics without grouping the entire anatomical cluster together.

</div>

<div style="width: 40%; float: left;">

<h3>Progress</h3>
Each case has undergone whole-brain tractography. Group coregistration, congealing, and clustering has taken place and the results have been confirmed as anatomically accurate by Lauren O'Donnell. The most relevant neuropsychological and clinical measures are being selected by David Tate. Correlations between these values and diffusion measures will be run on automatically outputted clusters to see if there are any differences between the 2 subject groups.

For analysis, we are first looking at the data in 2 ways: 1) Within the Normal Controls, we will be measuring FA/Mode/Trace/Axial Diffusion/Radial Diffusion for all the 200 bilateral tracts and correlating them with neuropsychological tests since trends would be expect in a healthy population. 2) Divide each of the 200 clusters down the midsagittal plane to separate the bilateral clusters into left and right sides. Then correlate with neuropsych.

</div>
</div>

<div style="width: 970%; float: left;">

==References==
* [http://lmi.bwh.harvard.edu/papers/papers/odonnellThesis06.html O'Donnell L. Cerebral White Matter Analysis Using Diffusion Imaging. Ph.D. thesis, Massachusetts Institute of Technology, 2006.]
* [http://pnl.bwh.harvard.edu/pub/papers_html/WhitfordMysell2009.html Whitford T, Kubicki M, King R, Khan U, Markant D, Alvarado J, McCarley R, Shenton ME. Abnormalities in tensor morphology in patients with schizophrenia: A dti study of the corpus callosum, 2009.]

</div>

2009 Summer Project Week FunctionalClusteringAnalysis

2009-06-22T17:26:20Z

Dougt:

__NOTOC__
<gallery>
Image:PW2009-v3.png|[[2009_Summer_Project_Week#Projects|Projects List]]
Image:Dougt_NAMIC1.jpg|Whole brain clustering in sagittal view.
Image:Dougt_NAMIC2.jpg|Whole brain clustering in axial view.
</gallery>

==Key Investigators==
* BWH: Doug Terry, Marek Kubicki, Andrew Rausch, Lauren O'Donnell, David Tate, CF Westin

<div style="margin: 20px;">
<div style="width: 27%; float: left; padding-right: 3%;">

<h3>Objective</h3>
We are using the clusters that are automatically outputted from whole-brain tractography and whole-brain congealing/clustering to draw correlations based on anatomical connectivity to see if there are group differences between chronic schizophrenics and controls.

</div>

<div style="width: 27%; float: left; padding-right: 3%;">

<h3>Approach, Plan</h3>

After whole brain tractography is completed for each case, the tracts for all the cases will be put through group coregistration/congealing. Then clustering takes place on the mean congealed brain. Specified number of clusters here is 200 for the whole brain, bilaterally. The clusters from the mean image are then transferred back to individual subject space. The mean FA/Mode/Trace will be taken from each cluster on each subject.

These diffusion values will then be correlated with clinical/neuropsych assessments. This is a way to look at possible deficits in schizophrenics without grouping the entire anatomical cluster together.

</div>

<div style="width: 40%; float: left;">

<h3>Progress</h3>
Each case has undergone whole-brain tractography. Group coregistration, congealing, and clustering has taken place and the results have been confirmed as anatomically accurate by Lauren O'Donnell. The most relevant neuropsychological and clinical measures are being selected by David Tate. Correlations between these values and diffusion measures will be run on automatically outputted clusters to see if there are any differences between the 2 subject groups.

For analysis, we are first looking at the data in 2 ways: 1) Within the Normal Controls, we will be measuring FA/Mode/Trace/Axial Diffusion/Radial Diffusion for all the 200 bilateral tracts and correlating them with neuropsychological tests since trends would be expect in a healthy population. 2) Divide each of the 200 clusters down the midsagittal plane to separate the bilateral clusters into left and right sides. Then correlate with neuropsych.

</div>
</div>

<div style="width: 970%; float: left;">

==References==
* [http://lmi.bwh.harvard.edu/papers/papers/odonnellThesis06.html O'Donnell L. Cerebral White Matter Analysis Using Diffusion Imaging. Ph.D. thesis, Massachusetts Institute of Technology, 2006.]
* [http://pnl.bwh.harvard.edu/pub/papers_html/WhitfordMysell2009.html Whitford T, Kubicki M, King R, Khan U, Markant D, Alvarado J, McCarley R, Shenton ME. Abnormalities in tensor morphology in patients with schizophrenia: A dti study of the corpus callosum, 2009.]

</div>

2009 Summer Project Week FunctionalClusteringAnalysis

2009-06-22T17:11:24Z

Dougt: /* Key Investigators */

__NOTOC__
<gallery>
Image:PW2009-v3.png|[[2009_Summer_Project_Week#Projects|Projects List]]
Image:Dougt_NAMIC1.jpg|Whole brain clustering in sagittal view.
Image:Dougt_NAMIC2.jpg|Whole brain clustering in axial view.
</gallery>

==Key Investigators==
* BWH: Doug Terry, Marek Kubicki, Andrew Rausch, Lauren O'Donnell, David Tate

<div style="margin: 20px;">
<div style="width: 27%; float: left; padding-right: 3%;">

<h3>Objective</h3>
We are using the clusters that are automatically outputted from whole-brain tractography and whole-brain congealing/clustering to draw correlations based on anatomical connectivity to see if there are group differences between chronic schizophrenics and controls.

</div>

<div style="width: 27%; float: left; padding-right: 3%;">

<h3>Approach, Plan</h3>

After whole brain tractography is completed for each case, the tracts for all the cases will be put through group coregistration/congealing. Then clustering takes place on the mean congealed brain. Specified number of clusters here is 200 for the whole brain, bilaterally. The clusters from the mean image are then transferred back to individual subject space. The mean FA/Mode/Trace will be taken from each cluster on each subject.

These diffusion values will then be correlated with clinical/neuropsych assessments. This is a way to look at possible deficits in schizophrenics without grouping the entire anatomical cluster together.

</div>

<div style="width: 40%; float: left;">

<h3>Progress</h3>
Each case has undergone whole-brain tractography. Group coregistration, congealing, and clustering has taken place and the results have been confirmed as anatomically accurate by Lauren O'Donnell. The most relevant neuropsychological and clinical measures are being selected by David Tate. Correlations between these values and diffusion measures will be run on automatically outputted clusters to see if there are any differences between the 2 subject groups.

For analysis, we are first looking at the data in 2 ways: 1) Within the Normal Controls, we will be measuring FA/Mode/Trace/Axial Diffusion/Radial Diffusion for all the 200 bilateral tracts and correlating them with neuropsychological tests since trends would be expect in a healthy population. 2) Divide each of the 200 clusters down the midsagittal plane to separate the bilateral clusters into left and right sides. Then correlate with neuropsych.

</div>
</div>

<div style="width: 970%; float: left;">

==References==
* [http://lmi.bwh.harvard.edu/papers/papers/odonnellThesis06.html O'Donnell L. Cerebral White Matter Analysis Using Diffusion Imaging. Ph.D. thesis, Massachusetts Institute of Technology, 2006.]
* [http://pnl.bwh.harvard.edu/pub/papers_html/WhitfordMysell2009.html Whitford T, Kubicki M, King R, Khan U, Markant D, Alvarado J, McCarley R, Shenton ME. Abnormalities in tensor morphology in patients with schizophrenia: A dti study of the corpus callosum, 2009.]

</div>

2009 Summer Project Week FunctionalClusteringAnalysis

2009-06-16T19:15:11Z

Dougt:

__NOTOC__
<gallery>
Image:PW2009-v3.png|[[2009_Summer_Project_Week|Project Week Main Page]]
Image:Dougt_NAMIC1.jpg|Whole brain clustering in sagittal view.
Image:Dougt_NAMIC2.jpg|Whole brain clustering in axial view.
</gallery>

==Key Investigators==
* BWH: Doug Terry, Marek Kubicki, Andrew Rausch, Lauren O'Donnell, David Tate

<div style="margin: 20px;">
<div style="width: 27%; float: left; padding-right: 3%;">

<h3>Objective</h3>
We are using the clusters that are automatically outputted from whole-brain tractography and whole-brain congealing/clustering to draw correlations based on anatomical connectivity to see if there are group differences between chronic schizophrenics and controls.

</div>

<div style="width: 27%; float: left; padding-right: 3%;">

<h3>Approach, Plan</h3>

After whole brain tractography is completed for each case, the tracts for all the cases will be put through group coregistration/congealing. Then clustering takes place on the mean congealed brain. Specified number of clusters here is 200 for the whole brain. The clusters from the mean image are then transferred back to individual subject space. The mean FA/Mode/Trace will be taken from each cluster on each subject.

These diffusion values will then be correlated with clinical/neuropsych assessments. This is a way to look at possible deficits in schizophrenics without grouping the entire anatomical cluster together.

</div>

<div style="width: 40%; float: left;">

<h3>Progress</h3>
Each case has undergone whole-brain tractography. Group coregistration, congealing, and clustering has also taken place, but there are issues with the validity of the group coregistration in the pipeline. Once this is worked out, the results from the clustering will be more reliable and the measures of diffusion (FA/Mode/Trace/Axial Diffusion/Radial Diffusion) will be extracted.

The most relevant neuropsychological and clinical measures are being selected by David Tate. Correlations between these values and diffusion measures will be run on automatically outputted clusters to see if there are any differences between the 2 subject groups.

</div>
</div>

<div style="width: 970%; float: left;">

==References==
* [http://lmi.bwh.harvard.edu/papers/papers/odonnellThesis06.html O'Donnell L. Cerebral White Matter Analysis Using Diffusion Imaging. Ph.D. thesis, Massachusetts Institute of Technology, 2006.]
* [http://pnl.bwh.harvard.edu/pub/papers_html/WhitfordMysell2009.html Whitford T, Kubicki M, King R, Khan U, Markant D, Alvarado J, McCarley R, Shenton ME. Abnormalities in tensor morphology in patients with schizophrenia: A dti study of the corpus callosum, 2009.]

</div>

File:Dougt NAMIC2.jpg

2009-06-16T19:12:32Z

Dougt:

File:Dougt NAMIC1.jpg

2009-06-16T19:10:59Z

Dougt:

2009 Summer Project Week

2009-06-16T19:10:20Z

Dougt:

Back to [[Project Events]], [[Events]]

[[Image:PW2009-v3.png|300px]]

*'''Dates:''' June 22-26, 2009
*'''Location:''' MIT. [[Meeting_Locations:MIT_Grier_A_%26B|Grier Rooms A & B: 34-401A & 34-401B]].

==Introduction to the FIRST JOINT PROJECT WEEK==

We are pleased to announce the FIRST JOINT PROJECT WEEK of hands-on research and development activity for Image-Guided Therapy and Neuroscience applications. Participants will engage in open source programming using the [[NA-MIC-Kit|NA-MIC Kit]], algorithm design, medical imaging sequence development, tracking experiments, and clinical application. The main goal of this event is to move forward the translational research deliverables of the sponsoring centers and their collaborators. Active and potential collaborators are encouraged and welcome to attend this event. This event will be set up to maximize informal interaction between participants.

Active preparation will begin on''' Thursday, April 16th at 3pm ET''', with a kick-off teleconference. Invitations to this call will be sent to members of the sponsoring communities, their collaborators, past attendees of the event, as well as any parties who have expressed an interest in working with these centers. The main goal of the kick-off call is to get an idea of which groups/projects will be active at the upcoming event, and to ensure that there is sufficient coverage for all. Subsequent teleconferences will allow for more focused discussions on individual projects and allow the hosts to finalize the project teams, consolidate any common components, and identify topics that should be discussed in breakout sessions. In the final days leading upto the meeting, all project teams will be asked to fill in a template page on this wiki that describes the objectives and plan of their projects.

The event itself will start off with a short presentation by each project team, driven using their previously created description, and will help all participants get acquainted with others who are doing similar work. In the rest of the week, about half the time will be spent in breakout discussions on topics of common interest of subsets of the attendees, and the other half will be spent in project teams, doing hands-on project work. The hands-on activities will be done in 30-50 small teams of size 2-4, each with a mix of multi-disciplinary expertise. To facilitate this work, a large room at MIT will be setup with several tables, with internet and power access, and each computer software development based team will gather on a table with their individual laptops, connect to the internet to download their software and data, and be able to work on their projects. Teams working on projects that require the use of medical devices will proceed to Brigham and Women's Hospital and carry out their experiments there. On the last day of the event, a closing presentation session will be held in which each project team will present a summary of what they accomplished during the week.

This event is part of the translational research efforts of [http://www.na-mic.org NA-MIC], [http://www.ncigt.org NCIGT], [http://nac.spl.harvard.edu/ NAC], [http://catalyst.harvard.edu/home.html Harvard Catalyst], and [http://www.cimit.org CIMIT]. It is an expansion of the NA-MIC Summer Project Week that has been held annually since 2005. It will be held every summer at MIT and Brigham and Womens Hospital in Boston, typically during the last full week of June, and in Salt Lake City in the winter, typically during the second week of January.

A summary of all past NA-MIC Project Events that this FIRST JOINT EVENT is based on is available [[Project_Events#Past|here]].

== Agenda==
* Monday
** noon-1pm lunch
**1pm: Welcome (Ron Kikinis)
** 1:05-3:30pm Introduce [[#Projects|Projects]] using templated wiki pages (all Project Leads) ([http://wiki.na-mic.org/Wiki/index.php/Project_Week/Template Wiki Template])
** 3:30-5:30pm Start project work
* Tuesday
** 8:30am breakfast
**9:30-10am: NA-MIC Kit Overview (Jim Miller)
** 10-10:30am Slicer 3.4 Update (Steve Pieper)
** 10:30-11am Slicer IGT and Imaging Kit Update Update (Noby Hata, Scott Hoge)
** 11am-12:00pm Breakout Session: [[2009 Project Week Breakout Session: Slicer-Python]] (Demian W)
** noon lunch
** 2:30pm-5pm: [[2009 Project Week Data Clinic|Data Clinic]] (Ron Kikinis)
** 5:30pm adjourn for day
* Wednesday
** 8:30am breakfast
** 9am-12pm Breakout Session: [[2009 Project Week Breakout Session: ITK]] (Luis Ibanez)
** noon lunch
** 2:30pm: Breakout Session: [[2009 Project Week Breakout Session: 3D+T Microscopy Cell Dataset Segmentation]] (Alex G.)
** 5:30pm adjourn for day
* Thursday
** 8:30am breakfast
** 9-11am [[Events:TutorialContestJune2009|Tutorial Contest Presentations]]
** noon lunch
** 2:30pm: Breakout Session: [[2009 Project Week Breakout Session: XNAT for Programmers]] (Dan M.)
** 5:30pm adjourn for day
* Friday
** 8:30am breakfast
** 10am-noon: [[Events:TutorialContestJune2009|Tutorial Contest Winner Announcement]] and [[#Projects|Project Progress Updates]]
*** Noon: Lunch boxes and adjourn by 1:30pm.
***We need to empty room by 1:30. You are welcome to use wireless in Stata.
***Please sign up for the developer [http://www.slicer.org/pages/Mailinglist mailing lists]
***Next Project Week [[AHM_2010|in Utah, January 4-8, 2010]]

== Projects ==

#[[2009_Summer_Project_Week_Slicer3_Cortical_Thickness_Pipeline|Cortical Thickness Pipeline]] (Clement Vachet UNC)
#[[2009_Summer_Project_Week_Prostate_Robotics |Prostate Robotics]] (Junichi Tokuda BWH)
#[[2009_Summer_Project_Week_Lupus_Lesion_Segmentation |Lupus Lesion Segmentation]] (Jeremy Bockholt MRN)
#[[Summer2009:VCFS| Pipeline development for VCFS]] (Marek Kubicki BWH)
#[[2009_Summer_Project_Week_Project_Segmentation_of_Muscoskeletal_Images]] (Harish Doddi Stanford)
#[[2009_Summer_Project_Week_Liver_Ablation_Slicer|Liver Ablation in Slicer]] (Ziv Yaniv Georgetown)
#[[Measuring Alcohol Stress Interaction]] (Vidya Rajgopalan Virginia Tech)
#[[2009_Summer_Project_Week_Skull_Stripping | Skull Stripping]] (Snehasish Roy JHU)
# [[MeshingSummer2009 | IAFE Mesh Modules - improvements and testing]] (Curt Lisle Knowledge Vis)
#[[2009_Summer_Project_Week_Slicer3_Adaptive_Radiotherapy|Adaptive Radiotherapy - Deformable registration and DICOMRT]] (Greg Sharp MGH)
#[[2009_Summer_Project_Week_Slicer3_Brainlab_Introduction|SLicer3, BioImage Suite and Brainlab - Introduction to UCLA]] (Haiying Liu BWH)
#[[2009_Summer_Project_Week_Slicer3_Brainlab_Demo|Demo Brainlab-BioImage Suite-Slicer in BWH OR]] (Haiying Liu BWH)
#[[2009_Summer_Project_Week_Multimodal_SPL_Brain_Atlas|Segmentation of thalamic nuclei from DTI]] (Ion-Florin Talos BWH)
#[[2009_Summer_Project_Week_Slicer3_Fibre_Dispersion|Slicer module for the computation of fibre dispersion and curving measures]] (Peter Savadjiev BWH)
#[[2009_Summer_Project_Week_Hageman_FMTractography | Fluid mechanics tractography and visualization]] (Nathan Hageman UCLA)
#[[2009_Summer_Project_Week_DWI_/_DTI_QC_and_Prepare_Tool:_DTIPrep | DWI/DTI QC and Preparation Tool: DTIPrep]] (Zhexing Liu UNC)
#[[2009_Summer_Project_Week_Hageman_DTIDigitalPhantom | DTI digital phantom generator to create validation data sets - webservice/cmdlin module/binaries are downloadable from UCLA ]] (Nathan Hageman UCLA)
# [[EPI Correction in Slicer3 | EPI Correction in Slicer3]] (Ran Tao Utah)
#[[2009_Summer_Project_Week_WML_SEgmentation |White Matter Lesion segmentation]] (Minjeong Kim UNC)
#[[2009_Summer_Project_Week-FastMarching_for_brain_tumor_segmentation |FastMarching for brain tumor segmentation]] (Andrey Fedorov BWH)
# [[EMSegment|EM Segment]] (Sylvain Jaume BWH)
#[[2009_Summer_Project_Week_Meningioma_growth_simulation|Meningioma growth simulation]] (Andrey Fedorov BWH)
#[[2009_Summer_Project_Week_Automatic_Brain_MRI_Pipeline|Automatic brain MRI processing pipeline]] (Marcel Prastawa Utah)
#[[2009_Summer_Project_Week_HAMMER_Registration | HAMMER Registration]] (Guorong Wu UNC)
#[[2009_Summer_Project_Week_Spherical_Mesh_Diffeomorphic_Demons_Registration |Spherical Mesh Diffeomorphic Demons Registration]] (Luis Ibanez Kitware)
# [[BSpline Registration in Slicer3 | BSpline Registration in Slicer3]] (Samuel Gerber Utah)
#[[2009_Summer_Project_Week_4D_Imaging| 4D Imaging (Perfusion, Cardiac, etc.) ]] (Junichi Tokuda BWH)
#[[2009_Summer_Project_Week_MRSI-Module|MRSI Module]] (Bjoern Menze MIT)
#[[2009_Summer_Project_Week_4D_Gated_US_In_Slicer |Gated 4D ultrasound reconstruction for Slicer3]] (Danielle Pace Robarts Institute)
# [[Integration of stereo video into Slicer3]] (Mehdi Esteghamatian Robarts Institute)
#[[2009_Summer_Project_Week_Statistical_Toolbox |multi-modality statistical toolbox for MR T1, T2, fMRI, DTI data]] (Diego Cantor Robarts Institute)
# [[Summer2009:Using_ITK_in_python| Using ITK in python]] (Steve Pieper BWH)
# [[Summer2009:Implementing_parallelism_in_python| Taking advantage of multicore machines & clusters with python]] (Julien de Siebenthal BWH)
# [[Summer2009:Using_client_server_paradigm_with_python_and_slicer| Deferring heavy computational tasks with Slicer python]] (Julien de Siebenthal BWH)
# [[Summer2009:Using_cython| Accelerating python with cython: application to stochastic tractography]] (Julien de Siebenthal BWH)
# [[2009_Summer_Project_Week_VTK_3D_Widgets_In_Slicer3|VTK 3d Widgets in Slicer3]] (Nicole Aucoin BWH)
# [[2009_Summer_Project_Week_Colors_Module |Updates to Slicer3 Colors module]] (Nicole Aucoin BWH)
# [[Plug-In 3D Viewer based on XIP|Plug-in 3D Viewer based on XIP]] (Lining Yang Siemens Research)
# [[Slicer3 Informatics Workflow Design & XNAT updates | Slicer3 Informatics Workflow Design & XNAT updates for Slicer]] (Wendy Plesniak BWH)
# [[Summer2009:Registration reproducibility in Slicer|Registration reproducibility in Slicer3]] (Andrey Fedorov BWH)
# [[Summer2009:The Vascular Modeling Toolkit in 3D Slicer | The Vascular Modeling Toolkit in 3D Slicer]] (Daniel Haehn BWH)
# [[Summer2009:Extension of the Command Line XML Syntax/Interface | Extension of the Command Line XML Syntax/Interface]] (Bennett Landman)
#[[2009_Summer_Project_Week_Slicer3_XNAT_UI | XNAT user interface improvements for NA-MIC]] (Dan Marcus WUSTL)
#[[2009_Summer_Project_Week_XNATFS | XNAT File System with FUSE]] (Dan Marcus WUSTL)
#[[2009_Summer_Project_Week_XNAT_i2b2|XNAT integration into Harvard Catalyst i2b2 framework]] (Yong Harvard)
#[[2009_Summer_Project_Week_Slicer3_registration| Slicer 3 registration ]] (Andrew Rausch)
#[[2009_Summer_Project_Week_Transrectal_Prostate_biopsy|Transrectal Prostate Biopsy]] (Andras Lasso Queen's)
#[[2009_Summer_Project_Week_3DGRASE|3D GRASE]] (Scott Hoge BWH)
#[[2009_Summer_Project_Week_TrigeminalNerve|Atlas to CT Registration in Trigeminal Neuralgia]] (Marta Peroni PoliMI, Maria Francesca Spadea UMG, Greg Sharp MGH)
#[[2009_Summer_Project_Week_FunctionalClusteringAnalysis|Functional Analysis of White Matter in Whole Brain Clustering of Schizophrenic Patients]] (Doug Terry, Marek Kubicki BWH)
#[[2009_Summer_Project_Week_Slicer|Integration of Flexible Surgical Instrument Modeling and Virtual Catheter with Slicer]] (Jayender Jagadeesan BWH)

===CUDA Projects===

This is a list of candidate cuda projects that will be discussed with Joe Stam shortly:

#[[2009_Summer_Project_Week_Registration_for_RT|2d/3d Registration (and GPGPU acceleration) for Radiation Therapy]] (Tina Kapur BWH)
#[[2009_Summer_Project_Week_Statistical_Toolbox |multi-modality statistical toolbox for MR T1, T2, fMRI, DTI data]] (Diego Cantor Robarts Institute)
#[[2009_Summer_Project_Week_Dose_Calculation |accelerate calculation for LDR seeds]] (Jack Blevins Acousticmed)
#[[2009_Summer_Project_Week_Cone_Beam_backprojection]](Zhou Shen U Michigan)
#[[2009_Summer_project_week_3d_Deformable_alignment]](Dan McShan U Michigan)
#[[Summer2009:Using_CUDA_for_stochastic_tractography|Developing interactive stochastic tractography using CUDA]] (Julien de Siebenthal BWH)
#acceleration of parallel real time processing of strain and elasticity images for monitoring of ablative therapy (Clif Burdette Acousticmed)

== Preparation ==

# Please make sure that you are on the http://public.kitware.com/cgi-bin/mailman/listinfo/na-mic-project-week mailing list
# Join the kickoff TCON on April 16, 3pm ET.
# [[Engineering:TCON_2009|June 18 TCON]] at 3pm ET to tie loose ends. Anyone with un-addressed questions should call.
# By 3pm ET on June 11, 2009: [[Project_Week/Template|Complete a templated wiki page for your project]]. Please do not edit the template page itself, but create a new page for your project and cut-and-paste the text from this template page. If you have questions, please send an email to tkapur at bwh.harvard.edu.
# By 3pm on June 18, 2009: Create a directory for each project on the [[Engineering:SandBox|NAMIC Sandbox]] (Zack)
## Commit on each sandbox directory the code examples/snippets that represent our first guesses of appropriate methods. (Luis and Steve will help with this, as needed)
## Gather test images in any of the Data sharing resources we have (e.g. the BIRN). These ones don't have to be many. At least three different cases, so we can get an idea of the modality-specific characteristics of these images. Put the IDs of these data sets on the wiki page. (the participants must do this.)
## Setup nightly tests on a separate Dashboard, where we will run the methods that we are experimenting with. The test should post result images and computation time. (Zack)
# Please note that by the time we get to the project event, we should be trying to close off a project milestone rather than starting to work on one...
# People doing Slicer related projects should come to project week with slicer built on your laptop.
## Projects to develop extension modules should work with the [http://viewvc.slicer.org/viewcvs.cgi/branches/Slicer-3-4/#dirlist Slicer-3-4 branch] (new code should not be checked into the branch).
## Projects to modify core behavior of slicer should be done on the [http://viewvc.slicer.org/viewcvs.cgi/trunk/ trunk].

==Attendee List==
If you plan to attend, please add your name here.

#Ron Kikinis, BWH (NA-MIC, NAC, NCIGT)
#Clare Tempany, BWH (NCIGT)
#Tina Kapur, BWH (NA-MIC, NCIGT)
#Steve Pieper, Isomics Inc
#Jim Miller, GE Research
#Xiaodong Tao, GE Research
#Randy Gollub, MGH
#Nicole Aucoin, BWH (NA-MIC) (Attending Tuesday-Friday)
#Dan Marcus, WUSTL
#Junichi Tokuda, BWH (NCIGT)
#Alex Gouaillard, Harvard Systems Biology
#Arnaud Gelas, Harvard Systems Biology
#Kishore Mosanliganti, Harvard Systems Biology
#Lydie Souhait, Harvard Systems Biology
#Luis Ibanez, Kitware Inc (Attending: Monday/Tuesday/Wednesday)
#Vincent Magnotta, UIowa
#Hans Johnson, UIowa
#Xenios Papademetris, Yale
#Gregory S. Fischer, WPI (Mon, Tue, Wed)
#Daniel Blezek, Mayo (Tue-Fri)
#Danielle Pace, Robarts Research Institute / UWO
#Clement Vachet, UNC-Chapel Hill
#Dave Welch, UIowa
#Demian Wassermann, Odyssée lab, INRIA, France
#Manasi Ramachandran, UIowa
#Greg Sharp, MGH
#Rui Li, MGH
#Mehdi Esteghamatian, Robarts Research Institute / UWO
#Misha Milchenko, WUSTL
#Kevin Archie, WUSTL
#Tim Olsen, WUSTL
#Wendy Plesniak BWH (NAC)
#Haiying Liu BWH (NCIGT)
#Curtis Lisle, KnowledgeVis / Isomics
#Diego Cantor, Robarts Research Institute / UWO
#Daniel Haehn, BWH
#Nicolas Rannou, BWH
#Sylvain Jaume, MIT
#Alex Yarmarkovich, Isomics
#Marco Ruiz, UCSD
#Andriy Fedorov, BWH (NA-MIC)
#Harish Doddi, Stanford University
#Scott Hoge, BWH (NCIGT)
#Vandana Mohan, Georgia Tech
#Ivan Kolosev, Georgia Tech
#Behnood Gholami, Georgia Tech
#James Balter, U Michigan
#Dan McShan, U Michigan
#Zhou Shen, U Michigan
#Maria Francesca Spadea, Italy
#Lining Yang, Siemens Corporate Research
#Beatriz Paniagua, UNC-Chapel Hill
#Bennett Landman, Johns Hopkins University
#Snehashis Roy, Johns Hopkins University
#Marta Peroni, Politecnico di Milano
#Sebastien Barre, Kitware, Inc.
#Samuel Gerber, SCI University of Utah
#Ran Tao, SCI University of Utah
#Marcel Prastawa, SCI University of Utah
#Katie Hayes, BWH (NA-MIC)
#Sonia Pujol, BWH (NA-MIC)
#Andras Lasso, Queen's University
#Yong Gao, MGH
#Minjeong Kim, UNC-Chapel Hill
#Guorong Wu, UNC-Chapel Hill
#Jeffrey Yager, UIowa
#Yanling Liu, SAIC/NCI-Frederick
#Ziv Yaniv, Georgetown
#Bjoern Menze, MIT
#Vidya Rajagopalan, Virginia Tech
#Sandy Wells, BWH (NAC, NCIGT)
#Lilla Zollei, MGH (NAC)
#Lauren O'Donnell, BWH
#Florin Talos, BWH (NAC)
#Nobuhiko Hata, BWH (NCIGT)
#Alark Joshi, Yale
#Yogesh Rathi, BWH
#Jimi Malcolm, BWH
#Dustin Scheinost, Yale
#Dominique Belhachemi, Yale
#Sam Song, JHU
#Nathan Cho, JHU
#Julien de Siebenthal, BWH
#Peter Savadjiev, BWH
#Carl-Fredrik Westin, BWH
#John Melonakos, AccelerEyes (Wed & Thu morning)
#Yi Gao, Georgia Tech
#Sylvain Bouix, BWH
#Zhexing Liu, UNC-CH
#Eric Melonakos, BWH
#Lei Qin, BWH
#Giovanna Danagoulian, BWH
#Andrew Rausch, BWH (Monday)
#Haytham Elhawary, BWH
#Jayender Jagadeesan, BWH
#Marek Kubicki, BWH
#Doug Terry, BWH
#Nathan Hageman, LONI (UCLA)
#Dana Peters, Beth Israel Deaconess
#Sun Woo Lee, BWH
# Melanie Grebe, Siemens Corporate Research
# Megumi Nakao, BWH/NAIST
# Moti Freiman, The Hebrew Univ. of Jerusalem
#Jack Blevins, Acoustic Med Systems
#Michael Halle, BWH
#Amanda Peters, Harvard SEAS
#Joe Stam, NVIDIA (Wednesday, Thursday)
#Petter Risholm, BWH (NCIGT)
#Kimberly Powell, NVIDIA (Wednesday)
#Padma Akella, BWH (NCIGT)
#Clif Burdette, Acousticmed (Mon, Tue, Wed)
#Mark Scully, MRN
#Jeremy Bockholt, MRN (tues-thurs)
#Curtis Rueden, UW-Madison
#Juhana Frosen, BWH (Tuesday)
#Andrzej Przybyszewski, UMass Medical School (Monday)
#Robert Yaffe, MGH
#Kenneth (Cal) Hisley, Des Moines University
#Ross Whitaker, University of Utah (Wed-Fri)

== Logistics ==
*'''Dates:''' June 22-26, 2009
*'''Location:''' MIT. [[Meeting_Locations:MIT_Grier_A_%26B|Grier Rooms A & B: 34-401A & 34-401B]].
*'''Registration Fee:''' $260 (covers the cost of breakfast, lunch and coffee breaks for the week). Due by Friday, June 12th, 2009. Please make checks out to "Massachusetts Institute of Technology" and mail to: Donna Kaufman, MIT, 77 Massachusetts Ave., 38-409a, Cambridge, MA 02139. Receipts will be provided by email as checks are received. Please send questions to dkauf at mit.edu. '''If this is your first event and you are attending for only one day, the registration fee is waived.''' Please let us know, so that we can cover the costs with one of our grants.
*'''Registration Method''' Add your name to the Attendee List section of this page
*'''Hotel:''' We have a group rate of $189/night (plus tax) at the Le Meridien (which used to be the Hotel at MIT). [http://www.starwoodmeeting.com/Book/MITDECSE Please click here to reserve.] This rate is good only through June 1.
*Here is some information about several other Boston area hotels that are convenient to NA-MIC events: [[Boston_Hotels|Boston_Hotels]]. Summer is tourist season in Boston, so please book your rooms early.
*2009 Summer Project Week [[NA-MIC/Projects/Theme/Template|'''Template''']]
*[[2008_Summer_Project_Week#Projects|Last Year's Projects as a reference]]
*For hosting projects, we are planning to make use of the NITRC resources. See [[NA-MIC_and_NITRC | Information about NITRC Collaboration]]

2009 Summer Project Week FunctionalClusteringAnalysis

2009-06-16T18:12:35Z

Dougt: Created page with '__NOTOC__ <gallery> Image:PW2009-v3.png|Project Week Main Page </gallery> ==Key Investigators== * BWH: Doug Terry, Marek Kubicki, Andrew Rausch, Lau...'

__NOTOC__
<gallery>
Image:PW2009-v3.png|[[2009_Summer_Project_Week|Project Week Main Page]]
</gallery>

==Key Investigators==
* BWH: Doug Terry, Marek Kubicki, Andrew Rausch, Lauren O'Donnell, David Tate

<div style="margin: 20px;">
<div style="width: 27%; float: left; padding-right: 3%;">

<h3>Objective</h3>
We are using the clusters that are automatically outputted from whole-brain tractography and whole-brain congealing/clustering to draw correlations based on anatomical connectivity to see if there are group differences between chronic schizophrenics and controls.

</div>

<div style="width: 27%; float: left; padding-right: 3%;">

<h3>Approach, Plan</h3>

After whole brain tractography is completed for each case, the tracts for all the cases will be put through group coregistration/congealing. Then clustering takes place on the mean congealed brain. Specified number of clusters here is 200 for the whole brain. The clusters from the mean image are then transferred back to individual subject space. The mean FA/Mode/Trace will be taken from each cluster on each subject.

These diffusion values will then be correlated with clinical/neuropsych assessments. This is a way to look at possible deficits in schizophrenics without grouping the entire anatomical cluster together.

</div>

<div style="width: 40%; float: left;">

<h3>Progress</h3>
Each case has undergone whole-brain tractography. Group coregistration, congealing, and clustering has also taken place, but there are issues with the validity of the group coregistration in the pipeline. Once this is worked out, the results from the clustering will be more reliable and the measures of diffusion (FA/Mode/Trace/Axial Diffusion/Radial Diffusion) will be extracted.

The most relevant neuropsychological and clinical measures are being selected by David Tate. Correlations between these values and diffusion measures will be run on automatically outputted clusters to see if there are any differences between the 2 subject groups.

</div>
</div>

<div style="width: 970%; float: left;">

==References==
* [http://lmi.bwh.harvard.edu/papers/papers/odonnellThesis06.html O'Donnell L. Cerebral White Matter Analysis Using Diffusion Imaging. Ph.D. thesis, Massachusetts Institute of Technology, 2006.]
* [http://pnl.bwh.harvard.edu/pub/papers_html/WhitfordMysell2009.html Whitford T, Kubicki M, King R, Khan U, Markant D, Alvarado J, McCarley R, Shenton ME. Abnormalities in tensor morphology in patients with schizophrenia: A dti study of the corpus callosum, 2009.]

</div>

2009 Summer Project Week

2009-06-16T14:37:45Z

Dougt:

Back to [[Project Events]], [[Events]]

[[Image:PW2009-v3.png|300px]]

*'''Dates:''' June 22-26, 2009
*'''Location:''' MIT. [[Meeting_Locations:MIT_Grier_A_%26B|Grier Rooms A & B: 34-401A & 34-401B]].

==Introduction to the FIRST JOINT PROJECT WEEK==

We are pleased to announce the FIRST JOINT PROJECT WEEK of hands-on research and development activity for Image-Guided Therapy and Neuroscience applications. Participants will engage in open source programming using the [[NA-MIC-Kit|NA-MIC Kit]], algorithm design, medical imaging sequence development, tracking experiments, and clinical application. The main goal of this event is to move forward the translational research deliverables of the sponsoring centers and their collaborators. Active and potential collaborators are encouraged and welcome to attend this event. This event will be set up to maximize informal interaction between participants.

Active preparation will begin on''' Thursday, April 16th at 3pm ET''', with a kick-off teleconference. Invitations to this call will be sent to members of the sponsoring communities, their collaborators, past attendees of the event, as well as any parties who have expressed an interest in working with these centers. The main goal of the kick-off call is to get an idea of which groups/projects will be active at the upcoming event, and to ensure that there is sufficient coverage for all. Subsequent teleconferences will allow for more focused discussions on individual projects and allow the hosts to finalize the project teams, consolidate any common components, and identify topics that should be discussed in breakout sessions. In the final days leading upto the meeting, all project teams will be asked to fill in a template page on this wiki that describes the objectives and plan of their projects.

The event itself will start off with a short presentation by each project team, driven using their previously created description, and will help all participants get acquainted with others who are doing similar work. In the rest of the week, about half the time will be spent in breakout discussions on topics of common interest of subsets of the attendees, and the other half will be spent in project teams, doing hands-on project work. The hands-on activities will be done in 30-50 small teams of size 2-4, each with a mix of multi-disciplinary expertise. To facilitate this work, a large room at MIT will be setup with several tables, with internet and power access, and each computer software development based team will gather on a table with their individual laptops, connect to the internet to download their software and data, and be able to work on their projects. Teams working on projects that require the use of medical devices will proceed to Brigham and Women's Hospital and carry out their experiments there. On the last day of the event, a closing presentation session will be held in which each project team will present a summary of what they accomplished during the week.

This event is part of the translational research efforts of [http://www.na-mic.org NA-MIC], [http://www.ncigt.org NCIGT], [http://nac.spl.harvard.edu/ NAC], [http://catalyst.harvard.edu/home.html Harvard Catalyst], and [http://www.cimit.org CIMIT]. It is an expansion of the NA-MIC Summer Project Week that has been held annually since 2005. It will be held every summer at MIT and Brigham and Womens Hospital in Boston, typically during the last full week of June, and in Salt Lake City in the winter, typically during the second week of January.

A summary of all past NA-MIC Project Events that this FIRST JOINT EVENT is based on is available [[Project_Events#Past|here]].

== Agenda==
* Monday
** noon-1pm lunch
**1pm: Welcome (Ron Kikinis)
** 1:05-3:30pm Introduce [[#Projects|Projects]] using templated wiki pages (all Project Leads) ([http://wiki.na-mic.org/Wiki/index.php/Project_Week/Template Wiki Template])
** 3:30-5:30pm Start project work
* Tuesday
** 8:30am breakfast
**9:30-10am: NA-MIC Kit Overview (Jim Miller)
** 10-10:30am Slicer 3.4 Update (Steve Pieper)
** 10:30-11am Slicer IGT and Imaging Kit Update Update (Noby Hata, Scott Hoge)
** 11am-12:00pm Breakout Session: [[2009 Project Week Breakout Session: Slicer-Python]] (Demian W)
** noon lunch
** 2:30pm-5pm: [[2009 Project Week Data Clinic|Data Clinic]] (Ron Kikinis)
** 5:30pm adjourn for day
* Wednesday
** 8:30am breakfast
** 9am-12pm Breakout Session: [[2009 Project Week Breakout Session: ITK]] (Luis Ibanez)
** noon lunch
** 2:30pm: Breakout Session: [[2009 Project Week Breakout Session: 3D+T Microscopy Cell Dataset Segmentation]] (Alex G.)
** 5:30pm adjourn for day
* Thursday
** 8:30am breakfast
** 9-11am [[Events:TutorialContestJune2009|Tutorial Contest Presentations]]
** noon lunch
** 2:30pm: Breakout Session: [[2009 Project Week Breakout Session: XNAT for Programmers]] (Dan M.)
** 5:30pm adjourn for day
* Friday
** 8:30am breakfast
** 10am-noon: [[Events:TutorialContestJune2009|Tutorial Contest Winner Announcement]] and [[#Projects|Project Progress Updates]]
*** Noon: Lunch boxes and adjourn by 1:30pm.
***We need to empty room by 1:30. You are welcome to use wireless in Stata.
***Please sign up for the developer [http://www.slicer.org/pages/Mailinglist mailing lists]
***Next Project Week [[AHM_2010|in Utah, January 4-8, 2010]]

== Projects ==

#[[2009_Summer_Project_Week_Slicer3_Cortical_Thickness_Pipeline|Cortical Thickness Pipeline]] (Clement Vachet UNC)
#[[2009_Summer_Project_Week_Prostate_Robotics |Prostate Robotics]] (Junichi Tokuda BWH)
#[[2009_Summer_Project_Week_Lupus_Lesion_Segmentation |Lupus Lesion Segmentation]] (Jeremy Bockholt MRN)
#[[Summer2009:VCFS| Pipeline development for VCFS]] (Marek Kubicki BWH)
#[[2009_Summer_Project_Week_Project_Segmentation_of_Muscoskeletal_Images]] (Harish Doddi Stanford)
#[[2009_Summer_Project_Week_Liver_Ablation_Slicer|Liver Ablation in Slicer]] (Ziv Yaniv Georgetown)
#[[Measuring Alcohol Stress Interaction]] (Vidya Rajgopalan Virginia Tech)
#[[2009_Summer_Project_Week_Skull_Stripping | Skull Stripping]] (Snehasish Roy JHU)
# [[MeshingSummer2009 | IAFE Mesh Modules - improvements and testing]] (Curt Lisle Knowledge Vis)
#[[2009_Summer_Project_Week_Slicer3_Adaptive_Radiotherapy|Adaptive Radiotherapy - Deformable registration and DICOMRT]] (Greg Sharp MGH)
#[[2009_Summer_Project_Week_Slicer3_Brainlab_Introduction|SLicer3, BioImage Suite and Brainlab - Introduction to UCLA]] (Haiying Liu BWH)
#[[2009_Summer_Project_Week_Slicer3_Brainlab_Demo|Demo Brainlab-BioImage Suite-Slicer in BWH OR]] (Haiying Liu BWH)
#[[2009_Summer_Project_Week_Multimodal_SPL_Brain_Atlas|Segmentation of thalamic nuclei from DTI]] (Ion-Florin Talos BWH)
#[[2009_Summer_Project_Week_Slicer3_Fibre_Dispersion|Slicer module for the computation of fibre dispersion and curving measures]] (Peter Savadjiev BWH)
#[[2009_Summer_Project_Week_Hageman_FMTractography | Fluid mechanics tractography and visualization]] (Nathan Hageman UCLA)
#[[2009_Summer_Project_Week_DWI_/_DTI_QC_and_Prepare_Tool:_DTIPrep | DWI/DTI QC and Preparation Tool: DTIPrep]] (Zhexing Liu UNC)
#[[2009_Summer_Project_Week_Hageman_DTIDigitalPhantom | DTI digital phantom generator to create validation data sets - webservice/cmdlin module/binaries are downloadable from UCLA ]] (Nathan Hageman UCLA)
# [[EPI Correction in Slicer3 | EPI Correction in Slicer3]] (Ran Tao Utah)
#[[2009_Summer_Project_Week_WML_SEgmentation |White Matter Lesion segmentation]] (Minjeong Kim UNC)
#[[2009_Summer_Project_Week-FastMarching_for_brain_tumor_segmentation |FastMarching for brain tumor segmentation]] (Andrey Fedorov BWH)
# [[EMSegment|EM Segment]] (Sylvain Jaume BWH)
#[[2009_Summer_Project_Week_Meningioma_growth_simulation|Meningioma growth simulation]] (Andrey Fedorov BWH)
#[[2009_Summer_Project_Week_Automatic_Brain_MRI_Pipeline|Automatic brain MRI processing pipeline]] (Marcel Prastawa Utah)
#[[2009_Summer_Project_Week_HAMMER_Registration | HAMMER Registration]] (Guorong Wu UNC)
#[[2009_Summer_Project_Week_Spherical_Mesh_Diffeomorphic_Demons_Registration |Spherical Mesh Diffeomorphic Demons Registration]] (Luis Ibanez Kitware)
# [[BSpline Registration in Slicer3 | BSpline Registration in Slicer3]] (Samuel Gerber Utah)
#[[2009_Summer_Project_Week_4D_Imaging| 4D Imaging (Perfusion, Cardiac, etc.) ]] (Junichi Tokuda BWH)
#[[2009_Summer_Project_Week_MRSI-Module|MRSI Module]] (Bjoern Menze MIT)
#[[2009_Summer_Project_Week_4D_Gated_US_In_Slicer |Gated 4D ultrasound reconstruction for Slicer3]] (Danielle Pace Robarts Institute)
# [[Integration of stereo video into Slicer3]] (Mehdi Esteghamatian Robarts Institute)
#[[2009_Summer_Project_Week_Statistical_Toolbox |multi-modality statistical toolbox for MR T1, T2, fMRI, DTI data]] (Diego Cantor Robarts Institute)
# [[Summer2009:Using_ITK_in_python| Using ITK in python]] (Steve Pieper BWH)
# [[Summer2009:Implementing_parallelism_in_python| Taking advantage of multicore machines & clusters with python]] (Julien de Siebenthal BWH)
# [[Summer2009:Using_client_server_paradigm_with_python_and_slicer| Deferring heavy computational tasks with Slicer python]] (Julien de Siebenthal BWH)
# [[Summer2009:Using_cython| Accelerating python with cython: application to stochastic tractography]] (Julien de Siebenthal BWH)
# [[2009_Summer_Project_Week_VTK_3D_Widgets_In_Slicer3|VTK 3d Widgets in Slicer3]] (Nicole Aucoin BWH)
# [[2009_Summer_Project_Week_Colors_Module |Updates to Slicer3 Colors module]] (Nicole Aucoin BWH)
# [[Plug-In 3D Viewer based on XIP|Plug-in 3D Viewer based on XIP]] (Lining Yang Siemens Research)
# [[Slicer3 Informatics Workflow Design & XNAT updates | Slicer3 Informatics Workflow Design & XNAT updates for Slicer]] (Wendy Plesniak BWH)
# [[Summer2009:Registration reproducibility in Slicer|Registration reproducibility in Slicer3]] (Andrey Fedorov BWH)
# [[Summer2009:The Vascular Modeling Toolkit in 3D Slicer | The Vascular Modeling Toolkit in 3D Slicer]] (Daniel Haehn BWH)
# [[Summer2009:Extension of the Command Line XML Syntax/Interface | Extension of the Command Line XML Syntax/Interface]] (Bennett Landman)
#[[2009_Summer_Project_Week_Slicer3_XNAT_UI | XNAT user interface improvements for NA-MIC]] (Dan Marcus WUSTL)
#[[2009_Summer_Project_Week_XNATFS | XNAT File System with FUSE]] (Dan Marcus WUSTL)
#[[2009_Summer_Project_Week_XNAT_i2b2|XNAT integration into Harvard Catalyst i2b2 framework]] (Yong Harvard)
#[[2009_Summer_Project_Week_Slicer3_registration| Slicer 3 registration ]] (Andrew Rausch)
#[[2009_Summer_Project_Week_Transrectal_Prostate_biopsy|Transrectal Prostate Biopsy]] (Andras Lasso Queen's)
#[[2009_Summer_Project_Week_3DGRASE|3D GRASE]] (Scott Hoge BWH)
#[[2009_Summer_Project_Week_TrigeminalNerve|Atlas to CT Registration in Trigeminal Neuralgia]] (Marta Peroni PoliMI, Maria Francesca Spadea UMG, Greg Sharp MGH)
#[[2009_Summer_Project_Week_FunctionalClusteringAnalysis|Functional Analysis of Whole Brain Clustering in Schizophrenia]] (Doug Terry, Marek Kubicki BWH)

===CUDA Projects===

This is a list of candidate cuda projects that will be discussed with Joe Stam shortly:

#[[2009_Summer_Project_Week_Registration_for_RT|2d/3d Registration (and GPGPU acceleration) for Radiation Therapy]] (Tina Kapur BWH)
#[[2009_Summer_Project_Week_Statistical_Toolbox |multi-modality statistical toolbox for MR T1, T2, fMRI, DTI data]] (Diego Cantor Robarts Institute)
#[[2009_Summer_Project_Week_Dose_Calculation |accelerate calculation for LDR seeds]] (Jack Blevins Acousticmed)
#[[2009_Summer_Project_Week_Cone_Beam_backprojection]](Zhou Shen U Michigan)
#[[2009_Summer_project_week_3d_Deformable_alignment]](Dan McShan U Michigan)
#[[Summer2009:Using_CUDA_for_stochastic_tractography|Developing interactive stochastic tractography using CUDA]] (Julien de Siebenthal BWH)
#acceleration of parallel real time processing of strain and elasticity images for monitoring of ablative therapy (Clif Burdette Acousticmed)

== Preparation ==

# Please make sure that you are on the http://public.kitware.com/cgi-bin/mailman/listinfo/na-mic-project-week mailing list
# Join the kickoff TCON on April 16, 3pm ET.
# [[Engineering:TCON_2009|June 18 TCON]] at 3pm ET to tie loose ends. Anyone with un-addressed questions should call.
# By 3pm ET on June 11, 2009: [[Project_Week/Template|Complete a templated wiki page for your project]]. Please do not edit the template page itself, but create a new page for your project and cut-and-paste the text from this template page. If you have questions, please send an email to tkapur at bwh.harvard.edu.
# By 3pm on June 18, 2009: Create a directory for each project on the [[Engineering:SandBox|NAMIC Sandbox]] (Zack)
## Commit on each sandbox directory the code examples/snippets that represent our first guesses of appropriate methods. (Luis and Steve will help with this, as needed)
## Gather test images in any of the Data sharing resources we have (e.g. the BIRN). These ones don't have to be many. At least three different cases, so we can get an idea of the modality-specific characteristics of these images. Put the IDs of these data sets on the wiki page. (the participants must do this.)
## Setup nightly tests on a separate Dashboard, where we will run the methods that we are experimenting with. The test should post result images and computation time. (Zack)
# Please note that by the time we get to the project event, we should be trying to close off a project milestone rather than starting to work on one...
# People doing Slicer related projects should come to project week with slicer built on your laptop.
## Projects to develop extension modules should work with the [http://viewvc.slicer.org/viewcvs.cgi/branches/Slicer-3-4/#dirlist Slicer-3-4 branch] (new code should not be checked into the branch).
## Projects to modify core behavior of slicer should be done on the [http://viewvc.slicer.org/viewcvs.cgi/trunk/ trunk].

==Attendee List==
If you plan to attend, please add your name here.

#Ron Kikinis, BWH (NA-MIC, NAC, NCIGT)
#Clare Tempany, BWH (NCIGT)
#Tina Kapur, BWH (NA-MIC, NCIGT)
#Steve Pieper, Isomics Inc
#Jim Miller, GE Research
#Xiaodong Tao, GE Research
#Randy Gollub, MGH
#Nicole Aucoin, BWH (NA-MIC) (Attending Tuesday-Friday)
#Dan Marcus, WUSTL
#Junichi Tokuda, BWH (NCIGT)
#Alex Gouaillard, Harvard Systems Biology
#Arnaud Gelas, Harvard Systems Biology
#Kishore Mosanliganti, Harvard Systems Biology
#Lydie Souhait, Harvard Systems Biology
#Luis Ibanez, Kitware Inc (Attending: Monday/Tuesday/Wednesday)
#Vincent Magnotta, UIowa
#Hans Johnson, UIowa
#Xenios Papademetris, Yale
#Gregory S. Fischer, WPI (Mon, Tue, Wed)
#Daniel Blezek, Mayo (Tue-Fri)
#Danielle Pace, Robarts Research Institute / UWO
#Clement Vachet, UNC-Chapel Hill
#Dave Welch, UIowa
#Demian Wassermann, Odyssée lab, INRIA, France
#Manasi Ramachandran, UIowa
#Greg Sharp, MGH
#Rui Li, MGH
#Mehdi Esteghamatian, Robarts Research Institute / UWO
#Misha Milchenko, WUSTL
#Kevin Archie, WUSTL
#Tim Olsen, WUSTL
#Wendy Plesniak BWH (NAC)
#Haiying Liu BWH (NCIGT)
#Curtis Lisle, KnowledgeVis / Isomics
#Diego Cantor, Robarts Research Institute / UWO
#Daniel Haehn, BWH
#Nicolas Rannou, BWH
#Sylvain Jaume, MIT
#Alex Yarmarkovich, Isomics
#Marco Ruiz, UCSD
#Andriy Fedorov, BWH (NA-MIC)
#Harish Doddi, Stanford University
#Scott Hoge, BWH (NCIGT)
#Vandana Mohan, Georgia Tech
#Ivan Kolosev, Georgia Tech
#Behnood Gholami, Georgia Tech
#James Balter, U Michigan
#Dan McShan, U Michigan
#Zhou Shen, U Michigan
#Maria Francesca Spadea, Italy
#Lining Yang, Siemens Corporate Research
#Beatriz Paniagua, UNC-Chapel Hill
#Bennett Landman, Johns Hopkins University
#Snehashis Roy, Johns Hopkins University
#Marta Peroni, Politecnico di Milano
#Sebastien Barre, Kitware, Inc.
#Samuel Gerber, SCI University of Utah
#Ran Tao, SCI University of Utah
#Marcel Prastawa, SCI University of Utah
#Katie Hayes, BWH (NA-MIC)
#Sonia Pujol, BWH (NA-MIC)
#Andras Lasso, Queen's University
#Yong Gao, MGH
#Minjeong Kim, UNC-Chapel Hill
#Guorong Wu, UNC-Chapel Hill
#Jeffrey Yager, UIowa
#Yanling Liu, SAIC/NCI-Frederick
#Ziv Yaniv, Georgetown
#Bjoern Menze, MIT
#Vidya Rajagopalan, Virginia Tech
#Sandy Wells, BWH (NAC, NCIGT)
#Lilla Zollei, MGH (NAC)
#Lauren O'Donnell, BWH
#Florin Talos, BWH (NAC)
#Nobuhiko Hata, BWH (NCIGT)
#Alark Joshi, Yale
#Yogesh Rathi, BWH
#Jimi Malcolm, BWH
#Dustin Scheinost, Yale
#Dominique Belhachemi, Yale
#Sam Song, JHU
#Nathan Cho, JHU
#Julien de Siebenthal, BWH
#Peter Savadjiev, BWH
#Carl-Fredrik Westin, BWH
#John Melonakos, AccelerEyes (Wed & Thu morning)
#Yi Gao, Georgia Tech
#Sylvain Bouix, BWH
#Zhexing Liu, UNC-CH
#Eric Melonakos, BWH
#Lei Qin, BWH
#Giovanna Danagoulian, BWH
#Andrew Rausch, BWH (Monday)
#Haytham Elhawary, BWH
#Jayender Jagadeesan, BWH
#Marek Kubicki, BWH
#Doug Terry, BWH
#Nathan Hageman, LONI (UCLA)
#Dana Peters, Beth Israel Deaconess
#Sun Woo Lee, BWH
# Melanie Grebe, Siemens Corporate Research
# Megumi Nakao, BWH/NAIST
# Moti Freiman, The Hebrew Univ. of Jerusalem
#Jack Blevins, Acoustic Med Systems
#Michael Halle, BWH
#Amanda Peters, Harvard SEAS
#Joe Stam, NVIDIA (Wednesday, Thursday)
#Petter Risholm, BWH (NCIGT)
#Kimberly Powell, NVIDIA (Wednesday)
#Padma Akella, BWH (NCIGT)
#Clif Burdette, Acousticmed (Mon, Tue, Wed)
#Mark Scully, MRN
#Jeremy Bockholt, MRN (tues-thurs)
#Curtis Rueden, UW-Madison
#Juhana Frosen, BWH (Tuesday)
#Andrzej Przybyszewski, UMass Medical School (Monday)
#Robert Yaffe, MGH
#Kenneth (Cal) Hisley, Des Moines University

== Logistics ==
*'''Dates:''' June 22-26, 2009
*'''Location:''' MIT. [[Meeting_Locations:MIT_Grier_A_%26B|Grier Rooms A & B: 34-401A & 34-401B]].
*'''Registration Fee:''' $260 (covers the cost of breakfast, lunch and coffee breaks for the week). Due by Friday, June 12th, 2009. Please make checks out to "Massachusetts Institute of Technology" and mail to: Donna Kaufman, MIT, 77 Massachusetts Ave., 38-409a, Cambridge, MA 02139. Receipts will be provided by email as checks are received. Please send questions to dkauf at mit.edu. '''If this is your first event and you are attending for only one day, the registration fee is waived.''' Please let us know, so that we can cover the costs with one of our grants.
*'''Registration Method''' Add your name to the Attendee List section of this page
*'''Hotel:''' We have a group rate of $189/night (plus tax) at the Le Meridien (which used to be the Hotel at MIT). [http://www.starwoodmeeting.com/Book/MITDECSE Please click here to reserve.] This rate is good only through June 1.
*Here is some information about several other Boston area hotels that are convenient to NA-MIC events: [[Boston_Hotels|Boston_Hotels]]. Summer is tourist season in Boston, so please book your rooms early.
*2009 Summer Project Week [[NA-MIC/Projects/Theme/Template|'''Template''']]
*[[2008_Summer_Project_Week#Projects|Last Year's Projects as a reference]]
*For hosting projects, we are planning to make use of the NITRC resources. See [[NA-MIC_and_NITRC | Information about NITRC Collaboration]]

2009 Summer Project Week

2009-06-04T14:41:33Z

Dougt:

Back to [[Project Events]], [[Events]]

*'''Dates:''' June 22-26, 2009
*'''Location:''' MIT. [[Meeting_Locations:MIT_Grier_A_%26B|Grier Rooms A & B: 34-401A & 34-401B]].

==Introduction to the FIRST JOINT PROJECT WEEK==

We are pleased to announce the FIRST JOINT PROJECT WEEK of hands-on research and development activity for Image-Guided Therapy and Neuroscience applications. Participants will engage in open source programming using the [[NA-MIC-Kit|NA-MIC Kit]], algorithm design, medical imaging sequence development, tracking experiments, and clinical application. The main goal of this event is to move forward the translational research deliverables of the sponsoring centers and their collaborators. Active and potential collaborators are encouraged and welcome to attend this event. This event will be set up to maximize informal interaction between participants.

Active preparation will begin on''' Thursday, April 16th at 3pm ET''', with a kick-off teleconference. Invitations to this call will be sent to members of the sponsoring communities, their collaborators, past attendees of the event, as well as any parties who have expressed an interest in working with these centers. The main goal of the kick-off call is to get an idea of which groups/projects will be active at the upcoming event, and to ensure that there is sufficient coverage for all. Subsequent teleconferences will allow for more focused discussions on individual projects and allow the hosts to finalize the project teams, consolidate any common components, and identify topics that should be discussed in breakout sessions. In the final days leading upto the meeting, all project teams will be asked to fill in a template page on this wiki that describes the objectives and plan of their projects.

The event itself will start off with a short presentation by each project team, driven using their previously created description, and will help all participants get acquainted with others who are doing similar work. In the rest of the week, about half the time will be spent in breakout discussions on topics of common interest of subsets of the attendees, and the other half will be spent in project teams, doing hands-on project work. The hands-on activities will be done in 30-50 small teams of size 2-4, each with a mix of multi-disciplinary expertise. To facilitate this work, a large room at MIT will be setup with several tables, with internet and power access, and each computer software development based team will gather on a table with their individual laptops, connect to the internet to download their software and data, and be able to work on their projects. Teams working on projects that require the use of medical devices will proceed to Brigham and Women's Hospital and carry out their experiments there. On the last day of the event, a closing presentation session will be held in which each project team will present a summary of what they accomplished during the week.

This event is part of the translational research efforts of [http://www.na-mic.org NA-MIC], [http://www.ncigt.org NCIGT], [http://nac.spl.harvard.edu/ NAC], [http://catalyst.harvard.edu/home.html Harvard Catalyst], and [http://www.cimit.org CIMIT]. It is an expansion of the NA-MIC Summer Project Week that has been held annually since 2005. It will be held every summer at MIT and Brigham and Womens Hospital in Boston, typically during the last full week of June, and in Salt Lake City in the winter, typically during the second week of January.

A summary of all past NA-MIC Project Events that this FIRST JOINT EVENT is based on is available [[Project_Events#Past|here]].

== Agenda==
* Monday
** noon-1pm lunch
**1pm: Welcome (Ron Kikinis)
** 1:05-3:30pm Introduce [[#Projects|Projects]] using templated wiki pages (all Project Leads) ([http://wiki.na-mic.org/Wiki/index.php/Project_Week/Template Wiki Template])
** 3:30-5:30pm Start project work
* Tuesday
** 8:30am breakfast
**9:30-10am: NA-MIC Kit Overview (Jim Miller)
** 10-10:30am Slicer 3.4 Update (Steve Pieper)
** 10:30-11am Slicer IGT and Imaging Kit Update Update (Noby Hata, Scott Hoge)
** 11am-12:00pm Breakout Session: [[2009 Project Week Breakout Session: Slicer-Python]] (Demian W)
** noon lunch
** 2:30pm-5pm: [[2009 Project Week Data Clinic|Data Clinic]] (Ron Kikinis)
** 5:30pm adjourn for day
* Wednesday
** 8:30am breakfast
** 9am-12pm Breakout Session: [[2009 Project Week Breakout Session: ITK]] (Luis Ibanez)
** noon lunch
** 2:30pm: Breakout Session: [[2009 Project Week Breakout Session: 3D+T Microscopy Cell Dataset Segmentation]] (Alex G.)
** 5:30pm adjourn for day
* Thursday
** 8:30am breakfast
** 9-11pm Tutorial Contest Presentations
** noon lunch
** 2:30pm: Breakout Session: [[2009 Project Week Breakout Session: XNAT]] (Dan M.)
** 5:30pm adjourn for day
* Friday
** 8:30am breakfast
** 10am-noon: [[Events:TutorialContestJune2009|Tutorial Contest Winner Announcement]] and [[#Projects|Project Progress Updates]]
*** Noon: Lunch boxes and adjourn by 1:30pm.
***We need to empty room by 1:30. You are welcome to use wireless in Stata.
***Please sign up for the developer [http://www.slicer.org/pages/Mailinglist mailing lists]
***Next Project Week [[AHM_2010|in Utah, January 4-8, 2010]]

== Projects ==

The list of projects for this week will go here.
=== Collaboration Projects ===
#[[2009_Summer_Project_Week_Project_Segmentation_of_Muscoskeletal_Images]]
#[[2009_Summer_Project_Week_4D_Imaging| 4D Imaging (Perfusion, Cardiac, etc.) ]] (Junichi, Dan Blezek?, Steve, Alex G?)
#[[2009_Summer_Project_Week_Liver_Ablation_Slicer|Liver Ablation in Slicer (Haiying, Ziv, Noby)]]
#[[2009_Summer_Project_Week_Slicer3_Brainlab_Introduction|SLicer3, BioImage Suite and Brainlab - Introduction to UCLA (Haiying, Xenios, Pratik, Nathan Hageman)]]
#Adaptive Radiotherapy - Deformable registration and DICOMRT (Greg Sharp, Steve, Wendy)
#Brain DTI Atlas? (Florin, Utah, UNC, GeorgiaTech)
#Slicer module for the computation of fibre dispersion and curving measures (Peter Savadjiev, C-F Westin)
#Xnat user interface improvements for NA-MIC (Dan M, Florin, Ron, Wendy)
#xnat and DICOMRT (Greg Sharp, Dan M) - might be done?
#Grid Wizard+xnat clinic (Clement Vachet)
#?Fluid Mechanincs Module (Nathan Hageman)
#?DTI digital phantom generator to create validation data sets - webservice/cmdlin module/binaries are downloadable from UCLA (Nathan Hageman)
#Cortical Thickness Pipeline (Clement Vachet, Ipek Oguz)
#[[2009_Summer_Project_Week_Slicer3_Brainlab_Demo|Demo Brainlab-BioImage Suite-Slicer in BWH OR (Haiying, Isaiah, Nathan Hageman)]]
#[[2009_Summer_Project_Week_Skull_Stripping | Skull Stripping]] (Xiaodong, Snehashis Roy)
#[[2009_Summer_Project_Week_HAMMER_Registration | HAMMER Registration]] (Guorong Wu, Xiaodong Tao, Jim Miller)
#[[2009_Summer_Project_Week_WML_SEgmentation |White Matter Lesion segmentation]] (Minjeong Kim, Xiaodong Tao, Jim Miller)
#[[2009_Summer_Project_Week-FastMarching_for_brain_tumor_segmentation |FastMarching for brain tumor segmentation]] (Fedorov, GeorgiaTech)
#[[2009_Summer_Project_Week_Meningioma_growth_simulation|Meningioma growth simulation]] (Fedorov, Marcel, Ron)
#Automatic brain MRI processing pipeline (Marcel, Hans)
#XNAT integration into Harvard Catalyst i2b2 framework(Gao, Yong)
#[[2009_Summer_Project_Week_Spherical_Mesh_Diffeomorphic_Demons_Registration |Spherical Mesh Diffeomorphic Demons Registration]] (Luis Ibanez,Thomas Yeo, Polina Goland), - (Mon, Tue, Wed)
#[[2009_Summer_Project_Week_MRSI-Module|MRSI Module]] (Bjoern Menze, Jeff Yager, Vince Magnotta)
#[[Measuring Alcohol Stress Interaction]] (Vidya Rajgopalan, Andrey Fedorov)
#DWI/DTI QC and Preparation Tool: DTIPrep (Zhexing Liu)

===IGT Projects:===
#[[2009_Summer_Project_Week_Prostate_Robotics |Prostate Robotics]] (Junichi, Sam, Nathan Cho, Jack), - Mon, Tue, Thursday 7pm-midnight)
#port 4d gated ultrasound code to Slicer - (Danielle)
#integration of stereo video into Slicer (Mehdi)
#multi-modality statistical toolbox for MR T1, T2, fMRI, DTI data (Diego, sylvain jaume, nicholas, noby)
#neuroendoscope workflow presentation (sebastien barre)
#breakout session on Dynamic Patient Models (James Balter)
#[[2009_Summer_Project_Week_Registration_for_RT|2d/3d Registration (and GPGPU acceleration) for Radiation Therapy]] (Sandy Wells, Jim Balter, and others)

===NA-MIC Engineering Projects===
# DICOM Validation and Cleanup Tool (Luis, Sid, Steve, Greg)
# [[Summer2009:Using_ITK_in_python| Using ITK in python]] (Steve, Demian, Jim)
# [[Summer2009:Implementing_parallelism_in_python| Taking advantage of multicore machines & clusters with python]] (Julien de Siebenthal, Sylvain Bouix)
# [[Summer2009:Using_client_server_paradigm_with_python_and_slicer| Deferring heavy computational tasks with python]] (Julien de Siebenthal, Sylvain Bouix)
# [[Summer2009:Using_CUDA_for_stochastic_tractography| Developing realtime feedback using CUDA]] (Julien de Siebenthal, Sylvain Bouix)
# [[2009_Summer_Project_Week_VTK_3D_Widgets_In_Slicer3|VTK 3d Widgets in Slicer3]] (Nicole, Karthik, Sebastien, Wendy)
# [[2009_Summer_Project_Week_Colors_Module |Updates to Slicer3 Colors module]] (Nicole)
# [[EM_Segmenter|EM Segmenter]] (Sylvain Jaume, Nicolas Rannou)
# Plug-in 3D Viewer based on XIP (Lining)
# [[MeshingSummer2009 | IAFE Mesh Modules - improvements and testing]] (Curt, Steve, Vince)
# [[Slicer3 Informatics Workflow Design & XNAT updates | Slicer3 Informatics Workflow Design & XNAT updates for Slicer]] (Wen, Steve, Dan M, Dan B)
# [[BSpline Registration in Slicer3 | BSpline Registration in Slicer3]] (Samuel Gerber,Jim Miller, Ross Whitaker)
# [[EPI Correction in Slicer3 | EPI Correction in Slicer3]] (Ran Tao, Jim Miller, Sylvain Bouix, Tom Fletcher, Ross Whitaker, Julien de Siebenthal)
# [[Summer2009:Registration reproducibility in Slicer|Registration reproducibility in Slicer3]] (Andriy, Luis, Bill, Jim, Steve)
# [[Summer2009:The Vascular Modeling Toolkit in 3D Slicer | The Vascular Modeling Toolkit in 3D Slicer]] (Daniel Haehn)

== Preparation ==

# Please make sure that you are on the http://public.kitware.com/cgi-bin/mailman/listinfo/na-mic-project-week mailing list
# Join the kickoff TCON on April 16, 3pm ET.
# [[Engineering:TCON_2009|June 18 TCON]] at 3pm ET to tie loose ends. Anyone with un-addressed questions should call.
# By 3pm ET on June 11, 2009: [[Project_Week/Template|Complete a templated wiki page for your project]]. Please do not edit the template page itself, but create a new page for your project and cut-and-paste the text from this template page. If you have questions, please send an email to tkapur at bwh.harvard.edu.
# By 3pm on June 18, 2009: Create a directory for each project on the [[Engineering:SandBox|NAMIC Sandbox]] (Zack)
## Commit on each sandbox directory the code examples/snippets that represent our first guesses of appropriate methods. (Luis and Steve will help with this, as needed)
## Gather test images in any of the Data sharing resources we have (e.g. the BIRN). These ones don't have to be many. At least three different cases, so we can get an idea of the modality-specific characteristics of these images. Put the IDs of these data sets on the wiki page. (the participants must do this.)
## Setup nightly tests on a separate Dashboard, where we will run the methods that we are experimenting with. The test should post result images and computation time. (Zack)
# Please note that by the time we get to the project event, we should be trying to close off a project milestone rather than starting to work on one...
# People doing Slicer related projects should come to project week with slicer built on your laptop.
## Projects to develop extension modules should work with the [http://viewvc.slicer.org/viewcvs.cgi/branches/Slicer-3-4/#dirlist Slicer-3-4 branch] (new code should not be checked into the branch).
## Projects to modify core behavior of slicer should be done on the [http://viewvc.slicer.org/viewcvs.cgi/trunk/ trunk].

==Attendee List==
If you plan to attend, please add your name here.

#Ron Kikinis, BWH (NA-MIC, NAC, NCIGT)
#Ferenc Jolesz, BWH (NCIGT, NAC)
#Clare Tempany, BWH (NCIGT)
#Tina Kapur, BWH (NA-MIC, NCIGT)
#Steve Pieper, Isomics Inc
#Jim Miller, GE Research
#Xiaodong Tao, GE Research
#Randy Gollub, MGH
#Nicole Aucoin, BWH (NA-MIC)
#Dan Marcus, WUSTL
#Junichi Tokuda, BWH (NCIGT)
#Alex Gouaillard, Harvard Systems Biology
#Arnaud Gelas, Harvard Systems Biology
#Kishore Mosanliganti, Harvard Systems Biology
#Lydie Souhait, Harvard Systems Biology
#Luis Ibanez, Kitware Inc
#Vincent Magnotta, UIowa
#Hans Johnson, UIowa
#Xenios Papademetris, Yale
#Gregory S. Fischer, WPI (Mon, Tue, Wed)
#Daniel Blezek, Mayo (Tue-Fri)
#Danielle Pace, Robarts Research Institute / UWO
#Clement Vachet, UNC-Chapel Hill
#Dave Welch, UIowa
#Demian Wassermann, Odyssée lab, INRIA, France
#Manasi Ramachandran, UIowa
#Greg Sharp, MGH
#Rui Li, MGH
#Mehdi Esteghamatian, Robarts Research Institute / UWO
#Misha Milchenko, WUSTL
#Kevin Archie, WUSTL
#Tim Olsen, WUSTL
#Wendy Plesniak BWH (NAC)
#Haiying Liu BWH (NCIGT)
#Curtis Lisle, KnowledgeVis / Isomics
#Diego Cantor, Robarts Research Institute / UWO
#Daniel Haehn, BWH
#Nicolas Rannou, BWH
#Sylvain Jaume, MIT
#Alex Yarmarkovich, Isomics
#Marco Ruiz, UCSD
#Andriy Fedorov, BWH (NA-MIC)
#Harish Doddi, Stanford University
#Saikat Pal, Stanford University
#Scott Hoge, BWH (NCIGT)
#Vandana Mohan, Georgia Tech
#Ivan Kolosev, Georgia Tech
#Behnood Gholami, Georgia Tech
#James Balter, U Michigan
#Dan McShan, U Michigan
#Zhou Shen, U Michigan
#Maria Francesca Spadea, Italy
#Lining Yang, Siemens Corporate Research
#Beatriz Paniagua, UNC-Chapel Hill
#Bennett Landman, Johns Hopkins University
#Snehashis Roy, Johns Hopkins University
#Marta Peroni, Politecnico di Milano
#Sebastien Barre, Kitware, Inc.
#Samuel Gerber, SCI University of Utah
#Ran Tao, SCI University of Utah
#Marcel Prastawa, SCI University of Utah
#Katie Hayes, BWH (NA-MIC)
#Sonia Pujol, BWH (NA-MIC)
#Andras Lasso, Queen's University
#Yong Gao, MGH
#Minjeong Kim, UNC-Chapel Hill
#Guorong Wu, UNC-Chapel Hill
#Jeffrey Yager, UIowa
#Yanling Liu, SAIC/NCI-Frederick
#Ziv Yaniv, Georgetown
#Bjoern Menze, MIT
#Vidya Rajagopalan, Virginia Tech
#Sandy Wells, BWH (NAC, NCIGT)
#Lilla Zollei, MGH (NAC)
#Lauren O'Donnell, BWH
#Florin Talos, BWH (NAC)
#Nobuhiko Hata, BWH (NCIGT)
#Alark Joshi, Yale
#Yogesh Rathi, BWH
#Jimi Malcolm, BWH
#Dustin Scheinost, Yale
#Dominique Belhachemi, Yale
#Sam Song, JHU
#Nathan Cho, JHU
#Julien de Siebenthal, BWH
#Peter Savadjiev, BWH
#Carl-Fredrik Westin, BWH
#John Melonakos, AccelerEyes (Wed & Thu morning)
#Yi Gao, Georgia Tech
#Sylvain Bouix, BWH
#Zhexing Liu, UNC-CH
#Eric Melonakos, BWH
#Lei Qin, BWH
#Giovanna Danagoulian, BWH
#Andrew Rausch, BWH (1st day only)
#Haytham Elhawary, BWH
#Jayender Jagadeesan, BWH
#Marek Kubicki, BWH
#Doug Terry, BWH

== Logistics ==
*'''Dates:''' June 22-26, 2009
*'''Location:''' MIT. [[Meeting_Locations:MIT_Grier_A_%26B|Grier Rooms A & B: 34-401A & 34-401B]].
*'''Registration Fee:''' $260 (covers the cost of breakfast, lunch and coffee breaks for the week). Due by Friday, June 12th, 2009. Please make checks out to "Massachusetts Institute of Technology" and mail to: Donna Kaufman, MIT, 77 Massachusetts Ave., 38-409a, Cambridge, MA 02139. Receipts will be provided by email as checks are received. Please send questions to dkauf at mit.edu. '''If this is your first event and you are attending for only one day, the registration fee is waived.''' Please let us know, so that we can cover the costs with one of our grants.
*'''Registration Method''' Add your name to the Attendee List section of this page
*'''Hotel:''' We have a group rate of $189/night (plus tax) at the Le Meridien (which used to be the Hotel at MIT). [http://www.starwoodmeeting.com/Book/MITDECSE Please click here to reserve.] This rate is good only through June 1.
*Here is some information about several other Boston area hotels that are convenient to NA-MIC events: [[Boston_Hotels|Boston_Hotels]]. Summer is tourist season in Boston, so please book your rooms early.
*2009 Summer Project Week [[NA-MIC/Projects/Theme/Template|'''Template''']]
*[[2008_Summer_Project_Week#Projects|Last Year's Projects as a reference]]
*For hosting projects, we are planning to make use of the NITRC resources. See [[NA-MIC_and_NITRC | Information about NITRC Collaboration]]

DBP2:Harvard:Brain Segmentation Roadmap

2009-06-03T18:31:02Z

Dougt:

Back to [[NA-MIC_Internal_Collaborations|NA-MIC Collaborations]], [[DBP2:Harvard|Harvard DBP 2]]
__NOTOC__
=Stochastic Tractography for VCFS=
== Roadmap ==

The main goal of this project is to develop end-to-end application that would be used to characterize anatomical connectivity abnormalities in the brain of patients with velocardiofacial syndrome (VCFS), and to link this information with deficits in schizophrenia. This page describes the technology roadmap for stochastic tractography, using newly acquired 3T data, NAMIC tools and slicer 3.

== Algorithm ==

[[Image:IC_sto_new.png|thumb|right|200px|<font size=1> Figure 1: Comparison of deterministic and stochastic tractography algorithms</font>]]
; A-Description
* Most tractography methods estimate fibers by tracing the maximum direction of diffusion. A limitation of this approach is that, in practice, several factors introduce uncertainty in the tracking procedure, including, noise, splitting and crossing fibers, head motion and image artifacts. To address this uncertainty, stochastic tractography methods have been developed to quantify the uncertainty associated with estimated fibers (Bjornemo et al., 2002). Method uses a propagation model based on stochastics and regularization, which allows paths originating at one point to branch and return a probability distribution of possible paths. The method utilizes principles of a statistical Monte Carlo method called Sequential Importance Sampling and Resampling (SISR). Based on probability functions, using a sequential importance sampling technique ([http://lmi.bwh.harvard.edu/papers/pdfs/2002/bjornemoMICCAI02.pdf Bjornemo et al., 2002]), one can generate thousands of fibers starting in the same point by sequentially drawing random step directions. This gives a very rich model of the fiber distribution, as contrasted with single fibers produced by conventional tractography methods. Moreover, from a large number of sampled paths, probability maps can be generated, providing better estimates of connectivity between several anatomical locations. A comparison of the algorithms can be seen here. (Figure 1)

[[Image:StochasticPic.PNG|thumb|right|200px|<font size=1>Figure 2: Stochastic tractography of uncinate fasciculis on anatomical data (left) and cingulum bungle on fMRI scan (right)</font>]]
; B-Possible Applications
* Since diffusion direction uncertainty within the gray matter is quite significant; principal diffusion direction approaches usually do not work for tracking between two gray matter regions. Thus if one requires finding connections between a priori selected anatomical gray matter regions, defined either by anatomical segmentations (in case of using structural ROI data), or functional activations (in case of megring DTI with fMRI), stochastic tractography seems to be the method of choice. Here is an example of this application to anatomical data (Figure 2, left image) and to fMRI data (Figure 2, right image).

* Stochastic Tractography is also comparable, if not better, in defining large white matter fiber bundles, especially those traveling through white matter regions characterized by increased diffusion uncertainty (fiber crossings). Example of such application to internal capsule. (Figure 3)

[[Image:IC-comp-new.png|thumb|right|200px|<font size=1>Figure 3: Streamline vs. stochastic tractography of the Internal Capsule</font>]]
; C-References

* [http://lmi.bwh.harvard.edu/papers/pdfs/2002/bjornemoMICCAI02.pdf Björnemo M, Brun A, Kikinis R, Westin CF. Regularized stochastic white matter tractography using diffusion tensor MRI. In Fifth International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI'02). Tokyo, Japan, 2002;435-442.]
* [http://lmi.bwh.harvard.edu/papers/pdfs/2006/frimanTMI06.pdf Friman, O., Farneback, G., Westin CF. A Bayesian Approach for Stochastic White Matter Tractography. IEEE Transactions on Medical Imaging, Vol 25, No. 8, Aug. 2006]

[[Image:StochasticGUI1.PNG|thumb|right|200px|<font size=1> Figure 4: Python Stochastic Tractography GUI </font>]]

==Module==
Can be found in: MODULES > PYTHON MODULES > PYTHON STOCHASTIC TRACTOGRAPHY
;Functionality of Python Stochastic Tractography module in Slicer 3.0
* IO:
Module reads files (DWI and ROIs) in nhdr format.
* Smoothing:
One can smooth the DWI data (only Gausian smoothing is supported at this time). We recommend it if the data is noisy.
* Brain Mask:
The Brain mask defines the volume in which the tensor will be computed and the tracts evaluated.
*Diffusion Tensor:
This step calculates the tensor and can output anisotropy indices (FA/Mode/Trace)
*Tractography:
Parameters that need to be adjusted:
:1. The amount of tracts that will be seeded from each voxel (we recommend between 500 and 1000 tracts, depending on the workstation power- 1000 tracts per voxel seeded within the large ROI for high resolution DWI can take a long time to compute).
:2. Maximum tract length (in mm), this can eliminate long, unwanted tracts if the regions for which connection is measured are located close to each other
:3. Step Size: distance between each re-estimation of tensors, usually between 0.5 and 1 mm
:4. Stopping criteria. This can be used on the top of WM mask, to terminate tracts (in case they really want to travel through the CSF, for example).
*Connectivity Map:
This step creates output probability maps.
:1. binary: each voxel is counted only once if at least one fiber pass through it
:2. cumulative: tracts are summed by voxel independently
:3. weighted: tracts are summed by voxel depending on their length
*Length Based:
This step will output only either the shortest 1/3, middle 1/3, or longest 1/3 of the tracts.
*Vicinity
This step traces back n number of steps from tract endpoint to check if track crosses target ROI. If so, tract is included.
*Threshold
This step will reject tracts whose endpoints are lower than the threshold value).
*Spherical ROI vicinity
This will make the ROI a sphere based on the ROI’s center of gravity (with the sphere’s radius being the distance from the center to the ROI’s furthest point). This sphere can be inflated by raising the Vicinity level to the number of steps you’d like to increase the ROI’s size by.

Then, probability maps can be saved as ROIs, and either used directly, or thresholded (at certain probability, step claimed by few publications to remove noise) in slicer to mask and compute average FA, Mode, Trace for entire connection. Diffusion indices can be also weighted by the probability of connection for each voxel.

:* Module documentation can be found here:
:**[[Media:IJdata.tar.gz|Training Dataset]]
:**[[Media:Stochastic_June09_1.ppt|Training Presentation]]
:**[[Media:Helix.zip|Sample Helix Dataset]]

==Work Accomplished==
[[Image:helix_withsmoothing.png|thumb|right|200px|<font size=1>Figure 6: Stochastic tractography from a single ROI on helix phantom</font>]]
; A - Optimization and testing of stochastic tractography algorithm :
:* Original methodological paper, as well as our first attempts to use the algorithm (CC+ and matlab scripts) have been done on old "NAMIC" 1.5T LSDI data ([https://portal.nbirn.net:443/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=gotoDirectory&CPq_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FHarvard%2Farchive_morph&up=CPq&JavaScript=enabled Structural MRI] and [https://portal.nbirn.net:443/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=gotoDirectory&up=7li&7li_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FHarvard%2Farchive_diffusion&JavaScript=enabled DTI data]).
:* Tri worked hard on making sure algorithm works on new high resolution 3T data (available here: [https://portal.nbirn.net/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=moveUpDir&gv7_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FPNL%2F3T_strct_dti_fmri%2Fcase01017&up=gv7&JavaScript=enabled|PNL 3T Data]).
:* Tests have been done also on the spiral diffusion phantom, to make sure diffusion directions and scanner coordinates are handled properly by the algorithm (Figure 6).

; B - Clinical Applications

:* Algorithm was used to trace and analyze anterior limb of the internal capsule on 1.5T data. It generated reacher representation of frontal fiber projections, it also turned out to be more sensitive to group differences in white matter integrity that conventional deterministic tractography (see Figure 6).
:* Algorithm was also used to trace smaller white matter fiber tracts, such as Cingulum, Fornix, Uncinate Fasciculus, Arcuate Fasciculus on 3T "Santa Fe" dataset (http://www.na-mic.org/Wiki/index.php/SanteFe.Tractography.Conference)

; C - References

:* [http://www.na-mic.org/Wiki/index.php/Image:IC_posternew.png Shenton, M.E., Ngo, T., Rosenberger, G., Westin, C.F., Levitt, J.J., McCarley, R.W., Kubicki, M. Study of Thalamo-Cortical White Matter Fiber Tract Projections in Schizophrenia Using Diffusion Stochastic Tractography. Poster presented at the 46th Meeting of the American College of Neuropsychopharmacology, Boca Raton, FL, December 2007.]
:* [http://www.na-mic.org/Wiki/index.php/Media:dougt_poster.pdf Terry DP, Rausch AC, Alvarado JL, Melonakos ED, Markant D, Westin CF, Kikinis R, von Siebenthal J, Shenton ME, Kubicki M. White Matter Properties of Emotion Related Connections in Schizophrenia. Poster presented at the 2009 Mysell Poster Day, Dept. of Psychiatry, Harvard Medical School, April 2009]
:* [http://www.na-mic.org/Wiki/index.php/Media:Jorge_poster.pdf Alvarado JL, Terry DP, Markant D, Ngo T, Kikinis R, Westin CF, McCarley RW, Shenton ME, Kubicki M. Study of Language-Related White Matter Tract Connections in Schizophrenia using Diffusion Stochastic Tractography. Poster presented at the 2009 Mysell Poster Day, Dept. of Psychiatry, Harvard Medical School, April 2009]
:* Melonakos ED, Shenton ME, Markant D, Alvarado J, Westin CF, Kubicki M. White Matter Properties of Orbitofrontal Connections in Schizophrenia. Poster being presented at the 64th Meeting of the Society of Biological Psychiatry. Vancouver, BC. May 2009.
:* Kubicki, M. Khan, U., Bobrow, L., O'Donnell, L. Pieper, S. Westin, CF., Shenton, ME. New Methods for Assessing Whole Brain DTI Abnormalities in Schizophrenia. Presentation given at the International Congress of World Psychiatric Association. Florence, Italy. April 2009.
:* Kubicki, M., Markant, D., Ngo, T., Westin, CF., McCarley, RW., Shenton, ME. Study of Language Related White Matter Fiber Tract Projections in Schizophrenia Using Diffusion Stochastic Tractography. Presentation given at the International Congress of World Psychiatric Association. Florence, Italy. April 2009.

== Work in Progress ==

; A - Optimization and Testing of stochastic tractography module :
:* Julien is testing now separate components of work flow, including the masks, and impact of their precision on stochastic output, as well as impact of number of seeding points on tractography results.
:* At the same time, we are testing the module on a Max Plank dataset of Anna Rotarska-Jagiela by looking at the connections through the corpus callosum of the left auditory cortex to the right auditory cortex as defined by fMRI activation maps as well as running this data for the tractography comparison project.
:* We are discussing possibility of adding project specific functionality to module, such as third ROI to guide tractography, nonlinear registration button for merging fMRI and DTI data.

; B - Related Clinical Projects

;* Arcuate Fasciculus Extraction Project
[[Image:STArcuate.jpg|thumb|right|200px|<font size=1>Figure 7: The arcuate fasciculus including seed, midpoint and target ROI's.</font>]]
We have started the project of investigating Arcuate Fasciculus using Stochastic Tractography (Figure 7). This structure is especially important in both VCFS and schizophrenia, as it connects language related areas (Brocka and Wernicke's), and is involved in language processing quite disturbed in schizophrenia patients. It also can not be reliably traced using deterministic tractography.
:Project involves:
:* Whole brain segmentation, and automatic extraction of regions interconnected by Arcuate Fasciculus (Inferior frontal and Superior Temporal Gyri).
:* White matter segmentation, in order to prevent algorithm from traveling through the ventricles, where diffusivity is high.
:* Non-linear registration of labelmaps to the DTI space.
:* Seeding tracts. We have piloted it using 5000 seeds per voxel, however it is quite time consuming running it on even most powerful computers in the lab, so we have experimented with smaller number of seeds per voxel. We tested 1000 seeds, which gave virtually identical results.
:* Extracting path of interest, and calculating FA along the path for group comparison. Presentation of previous results for 7 schizophrenics and 12 control subjects, can be found here: [[Media:NAMIC_AHM_Arcuate.ppt|Progress Report Presentation]].
:* Results presentation and paper submission. Abstract was submitted and accepted for presentation at World Biological Psychiatry Symposium in Venice, Italy. Paper is in preparation.

;* Semantic Network Connectivity Project
We use combination of fMRI and DTI data to define and characterize functional and anatomical connectivity within the semantic processing network in schizophrenia.
:Project involves:
:* fMRI data analysis and identification of functional nodes involved in semantic processing in healthy controls and seubjects with schizophrenia
:* Analysis of functional connectivity (using FSL) between nodes of semantic network
:* Whole brain Voxel Based analysis of DTI data in same population
:* Use of stochastic tractography to identify connections between functional nodes
:* Correlational analysis involving anatomical and functional connectivity data.
:* Results presentation and paper submission. Data was presented at HBM conference in 2007, paper has been submitted to HBM.
[[Image:Anna.png|thumb|right|200px|<font size=1>Figure 8: Max Plank data showing tracts through the corpus connecting 2 cortical ROIs defined by fMRI activations.</font>]]

;* Study of Default Network
We have started collaboration with Department of Neurophysiology Max Planck Institute in Frankfurt (Anna Rotarska contact person). They have dataset containing DTI and resting state fMRI in schizophrenia, and want to use stochastic tractography to measure integrity of anatomical connections within the default network in schizophrenia. Their fMRI data has been co-registered with anatomical scans, and has been put it into DTI space. Figure 8 shows pilot data of the white matter connections through the corpus callosum between the left and the right auditory cortex ROIs as defined by fMRI data. This data will be also used to test robustness of our module, since data was collected on a different scanner (3T Philips).

;* Tractography Comparison Project
We are also working on a [http://www.na-mic.org/Wiki/index.php/SanteFe.Tractography.Conference tractography comparison project]dataset, where we apply stochastic tractography to phantom, as well as test dataset.

===Staffing Plan===

* Sylvain and Yogesh are the DBP resources charged with adapting the tools in the NA-MIC Kit to the DBP needs
* Doug Markant, our NAMIC RA has left the lab, and now Doug Terry, is a new NAMIC RA.
* Julien is our new NAMIC software engineer. He is responsible for improving the STM (Stochastic Tractography Module), and making sure software works with STM compliant datasets.
[[Link Progress| Development Progress]]
* Polina is the algorithm core contact
* Brad is the engineering core contact

===Schedule===

* '''10/2007''' - Optimization of Stochastic Tractography algorythm for 1.5T data.
* '''10/2007''' - Algorythm testing on Santa Fe data set and diffusion phantom.
* '''06/2008''' - Optimization of Stochastic Tractography algorythm for 3T data.
* '''11/2008''' - Slicer 3 module prototype using python.
* '''12/2008''' - Slicer 3 module official release
* '''12/2008''' - Documentation and packaging for dissemination.
* '''12/2008''' - Arcuate Fasciculus results.
* '''01/2009''' - Arcuate Fasciuclus first draft of the paper.
* '''05/2009''' - Distortion correction and nonlinear registration added to the module
* '''05/2009''' - Symposium on tractography, including stochastic methods at World Biological Psychiatry Symposium in Florence, Italy.
* '''05/2009''' - Presentation of Arcuate Fasciculus findings at World Biological Psychiatry Symposium in Florence, Italy.

===Team and Institute===
*PI: Marek Kubicki (kubicki at bwh.harvard.edu)
*DBP2 Investigators: Sylvain Bouix, Yogesh Rathi, Julien de Siebenthal
*NA-MIC Engineering Contact: Brad Davis, Kitware
*NA-MIC Algorithms Contact: Polina Gollard, MIT

===Publications===

''In print''

*[http://www.na-mic.org/pages/Special:Publications?text=Kubicki+AND+Westin+AND+DTI&submit=Search&words=all&title=checked&keywords=checked&authors=checked&abstract=checked&sponsors=checked&searchbytag=checked| NA-MIC Publications Database - Clinical Applications]

*[http://www.na-mic.org/pages/Special:Publications?text=Ngo+AND+Golland&submit=Search&words=all&title=checked&keywords=checked&authors=checked&abstract=checked&sponsors=checked&searchbytag=checked| NA-MIC Publications Database - Algorithms Development]

[[Category: Schizophrenia]] [[Category: Diffusion MRI]] [[Category: Segmentation]]

Python Stochastic Tractography Tutorial

2009-06-03T18:29:27Z

Dougt:

{|
|[[Image:NAMIC-SLC.jpg|thumb|320px|Return to [[2009_Winter_Project_Week|Project Week Main Page]] ]]
|[[Image:Cc_01.png|thumb|340px|Model of the Corpus Callosum from Stochastic Tractography (coronal) ]]
|[[Image:Cc_02.png|thumb|340px|Model of the Corpus Callosum from Stochastic Tractography (sagittal) ]]
|[[Image:helix_withsmoothing.png|thumb|220px|Stochastic Tractography on Phantom]]
|}

__NOTOC__

===Key Investigators===
*PI: Marek Kubicki (kubicki at bwh.harvard.edu)
*DBP2 Investigators: Sylvain Bouix, Yogesh Rathi, Julien de Siebenthal, RA Doug Terry
*NA-MIC Engineering Contact: Brad Davis, Kitware
*NA-MIC Algorithms Contact: Polina Gollard, MIT

Go to [http://www.slicer.org/slicerWiki/index.php/Modules:StochasticTractography-Documentation-3.4 Slicer 3.4 module documentation] for more information

===Tutorial===
*[[Media:IJdata.tar.gz|Training Dataset]]
*[[Media:Stochastic_June09_1.ppt|Training Presentation]]
*[[Media:Helix.zip|Sample Helix Dataset]]

__NOTOC__
<div style="margin: 20px;">
<div style="width: 40%; float: left; padding-right: 3%;">
<h1>Objective</h1>
To address the uncertainty of deterministic tractography methods, stochastic tractography methods have been developed to quantify the uncertainty associated with estimated fibers (Bjornemo et al., 2002). Method uses a propagation model based on stochastics and regularization, which allows paths originating at one point to branch and return a probability distribution of possible paths. One can generate thousands of fibers starting in the same point by sequentially drawing random step directions. This gives a very rich model of the fiber distribution, as contrasted with single fibers produced by conventional tractography methods. Moreover, from a large number of sampled paths, probability maps can be generated, providing better estimates of connectivity between several anatomical locations.
</div>

<div style="width: 27%; float: left; padding-right: 3%;">
<h1>Approach, Plan</h1>
This tutorial will walk users through the multiple stages of the python stochastic tractography module, where the inputs are a DWI and a region of interest labelmap. This will include:
# Loading the Volumes
# Extracting DWI and getting Baseline DWI threshold levels
# Using features of the module including smoothing, brain mask generation, and other settings.
# Generating a Connectivity Map, which is able to be thresholded and customized.
</div>

<div style="width: 27%; float: left;">
<h1>Progress</h1>
*A tutorial has been developed and data is available for tutorial.
*We have started the project of investigating Arcuate Fasciculus using Stochastic Tractography. This structure is especially important in both VCFS and schizophrenia, as it connects language related areas (Brocka and Wernicke's), and is involved in language processing quite disturbed in schizophrenia patients. It also can not be reliably traced using deterministic tractography.
</div>

<br style="clear: both;" />
</div>

File:Stochastic June09 1.ppt

2009-06-03T18:29:05Z

Dougt:

DBP2:Harvard:Brain Segmentation Roadmap

2009-06-03T18:27:59Z

Dougt: /* Module */

Back to [[NA-MIC_Internal_Collaborations|NA-MIC Collaborations]], [[DBP2:Harvard|Harvard DBP 2]]
__NOTOC__
=Stochastic Tractography for VCFS=
== Roadmap ==

The main goal of this project is to develop end-to-end application that would be used to characterize anatomical connectivity abnormalities in the brain of patients with velocardiofacial syndrome (VCFS), and to link this information with deficits in schizophrenia. This page describes the technology roadmap for stochastic tractography, using newly acquired 3T data, NAMIC tools and slicer 3.

== Algorithm ==

[[Image:IC_sto_new.png|thumb|right|200px|<font size=1> Figure 1: Comparison of deterministic and stochastic tractography algorithms</font>]]
; A-Description
* Most tractography methods estimate fibers by tracing the maximum direction of diffusion. A limitation of this approach is that, in practice, several factors introduce uncertainty in the tracking procedure, including, noise, splitting and crossing fibers, head motion and image artifacts. To address this uncertainty, stochastic tractography methods have been developed to quantify the uncertainty associated with estimated fibers (Bjornemo et al., 2002). Method uses a propagation model based on stochastics and regularization, which allows paths originating at one point to branch and return a probability distribution of possible paths. The method utilizes principles of a statistical Monte Carlo method called Sequential Importance Sampling and Resampling (SISR). Based on probability functions, using a sequential importance sampling technique ([http://lmi.bwh.harvard.edu/papers/pdfs/2002/bjornemoMICCAI02.pdf Bjornemo et al., 2002]), one can generate thousands of fibers starting in the same point by sequentially drawing random step directions. This gives a very rich model of the fiber distribution, as contrasted with single fibers produced by conventional tractography methods. Moreover, from a large number of sampled paths, probability maps can be generated, providing better estimates of connectivity between several anatomical locations. A comparison of the algorithms can be seen here. (Figure 1)

[[Image:StochasticPic.PNG|thumb|right|200px|<font size=1>Figure 2: Stochastic tractography of uncinate fasciculis on anatomical data (left) and cingulum bungle on fMRI scan (right)</font>]]
; B-Possible Applications
* Since diffusion direction uncertainty within the gray matter is quite significant; principal diffusion direction approaches usually do not work for tracking between two gray matter regions. Thus if one requires finding connections between a priori selected anatomical gray matter regions, defined either by anatomical segmentations (in case of using structural ROI data), or functional activations (in case of megring DTI with fMRI), stochastic tractography seems to be the method of choice. Here is an example of this application to anatomical data (Figure 2, left image) and to fMRI data (Figure 2, right image).

* Stochastic Tractography is also comparable, if not better, in defining large white matter fiber bundles, especially those traveling through white matter regions characterized by increased diffusion uncertainty (fiber crossings). Example of such application to internal capsule. (Figure 3)

[[Image:IC-comp-new.png|thumb|right|200px|<font size=1>Figure 3: Streamline vs. stochastic tractography of the Internal Capsule</font>]]
; C-References

* [http://lmi.bwh.harvard.edu/papers/pdfs/2002/bjornemoMICCAI02.pdf Björnemo M, Brun A, Kikinis R, Westin CF. Regularized stochastic white matter tractography using diffusion tensor MRI. In Fifth International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI'02). Tokyo, Japan, 2002;435-442.]
* [http://lmi.bwh.harvard.edu/papers/pdfs/2006/frimanTMI06.pdf Friman, O., Farneback, G., Westin CF. A Bayesian Approach for Stochastic White Matter Tractography. IEEE Transactions on Medical Imaging, Vol 25, No. 8, Aug. 2006]

[[Image:StochasticGUI1.PNG|thumb|right|200px|<font size=1> Figure 4: Python Stochastic Tractography GUI </font>]]

==Module==
Can be found in: MODULES > PYTHON MODULES > PYTHON STOCHASTIC TRACTOGRAPHY
;Functionality of Python Stochastic Tractography module in Slicer 3.0
* IO:
Module reads files (DWI and ROIs) in nhdr format.
* Smoothing:
One can smooth the DWI data (only Gausian smoothing is supported at this time). We recommend it if the data is noisy.
* Brain Mask:
The Brain mask defines the volume in which the tensor will be computed and the tracts evaluated.
*Diffusion Tensor:
This step calculates the tensor and can output anisotropy indices (FA/Mode/Trace)
*Tractography:
Parameters that need to be adjusted:
:1. The amount of tracts that will be seeded from each voxel (we recommend between 500 and 1000 tracts, depending on the workstation power- 1000 tracts per voxel seeded within the large ROI for high resolution DWI can take a long time to compute).
:2. Maximum tract length (in mm), this can eliminate long, unwanted tracts if the regions for which connection is measured are located close to each other
:3. Step Size: distance between each re-estimation of tensors, usually between 0.5 and 1 mm
:4. Stopping criteria. This can be used on the top of WM mask, to terminate tracts (in case they really want to travel through the CSF, for example).
*Connectivity Map:
This step creates output probability maps.
:1. binary: each voxel is counted only once if at least one fiber pass through it
:2. cumulative: tracts are summed by voxel independently
:3. weighted: tracts are summed by voxel depending on their length
*Length Based:
This step will output only either the shortest 1/3, middle 1/3, or longest 1/3 of the tracts.
*Vicinity
This step traces back n number of steps from tract endpoint to check if track crosses target ROI. If so, tract is included.
*Threshold
This step will reject tracts whose endpoints are lower than the threshold value).
*Spherical ROI vicinity
This will make the ROI a sphere based on the ROI’s center of gravity (with the sphere’s radius being the distance from the center to the ROI’s furthest point). This sphere can be inflated by raising the Vicinity level to the number of steps you’d like to increase the ROI’s size by.

Then, probability maps can be saved as ROIs, and either used directly, or thresholded (at certain probability, step claimed by few publications to remove noise) in slicer to mask and compute average FA, Mode, Trace for entire connection. Diffusion indices can be also weighted by the probability of connection for each voxel.

:* Module documentation can be found here:
:**[[Media:IJdata.tar.gz|Training Dataset]]
:**[[Media:Stochastic_June09.ppt|Training Presentation]]
:**[[Media:Helix.zip|Sample Helix Dataset]]

==Work Accomplished==
[[Image:helix_withsmoothing.png|thumb|right|200px|<font size=1>Figure 6: Stochastic tractography from a single ROI on helix phantom</font>]]
; A - Optimization and testing of stochastic tractography algorithm :
:* Original methodological paper, as well as our first attempts to use the algorithm (CC+ and matlab scripts) have been done on old "NAMIC" 1.5T LSDI data ([https://portal.nbirn.net:443/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=gotoDirectory&CPq_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FHarvard%2Farchive_morph&up=CPq&JavaScript=enabled Structural MRI] and [https://portal.nbirn.net:443/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=gotoDirectory&up=7li&7li_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FHarvard%2Farchive_diffusion&JavaScript=enabled DTI data]).
:* Tri worked hard on making sure algorithm works on new high resolution 3T data (available here: [https://portal.nbirn.net/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=moveUpDir&gv7_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FPNL%2F3T_strct_dti_fmri%2Fcase01017&up=gv7&JavaScript=enabled|PNL 3T Data]).
:* Tests have been done also on the spiral diffusion phantom, to make sure diffusion directions and scanner coordinates are handled properly by the algorithm (Figure 6).

; B - Clinical Applications

:* Algorithm was used to trace and analyze anterior limb of the internal capsule on 1.5T data. It generated reacher representation of frontal fiber projections, it also turned out to be more sensitive to group differences in white matter integrity that conventional deterministic tractography (see Figure 6).
:* Algorithm was also used to trace smaller white matter fiber tracts, such as Cingulum, Fornix, Uncinate Fasciculus, Arcuate Fasciculus on 3T "Santa Fe" dataset (http://www.na-mic.org/Wiki/index.php/SanteFe.Tractography.Conference)

; C - References

:* [http://www.na-mic.org/Wiki/index.php/Image:IC_posternew.png Shenton, M.E., Ngo, T., Rosenberger, G., Westin, C.F., Levitt, J.J., McCarley, R.W., Kubicki, M. Study of Thalamo-Cortical White Matter Fiber Tract Projections in Schizophrenia Using Diffusion Stochastic Tractography. Poster presented at the 46th Meeting of the American College of Neuropsychopharmacology, Boca Raton, FL, December 2007.]
:* [http://www.na-mic.org/Wiki/index.php/Media:dougt_poster.pdf Terry DP, Rausch AC, Alvarado JL, Melonakos ED, Markant D, Westin CF, Kikinis R, von Siebenthal J, Shenton ME, Kubicki M. White Matter Properties of Emotion Related Connections in Schizophrenia. Poster presented at the 2009 Mysell Poster Day, Dept. of Psychiatry, Harvard Medical School, April 2009]
:* [http://www.na-mic.org/Wiki/index.php/Media:Jorge_poster.pdf Alvarado JL, Terry DP, Markant D, Ngo T, Kikinis R, Westin CF, McCarley RW, Shenton ME, Kubicki M. Study of Language-Related White Matter Tract Connections in Schizophrenia using Diffusion Stochastic Tractography. Poster presented at the 2009 Mysell Poster Day, Dept. of Psychiatry, Harvard Medical School, April 2009]
:* Melonakos ED, Shenton ME, Markant D, Alvarado J, Westin CF, Kubicki M. White Matter Properties of Orbitofrontal Connections in Schizophrenia. Poster being presented at the 64th Meeting of the Society of Biological Psychiatry. Vancouver, BC. May 2009.
:* Kubicki, M. Khan, U., Bobrow, L., O'Donnell, L. Pieper, S. Westin, CF., Shenton, ME. New Methods for Assessing Whole Brain DTI Abnormalities in Schizophrenia. Presentation given at the International Congress of World Psychiatric Association. Florence, Italy. April 2009.
:* Kubicki, M., Markant, D., Ngo, T., Westin, CF., McCarley, RW., Shenton, ME. Study of Language Related White Matter Fiber Tract Projections in Schizophrenia Using Diffusion Stochastic Tractography. Presentation given at the International Congress of World Psychiatric Association. Florence, Italy. April 2009.

== Work in Progress ==

; A - Optimization and Testing of stochastic tractography module :
:* Julien is testing now separate components of work flow, including the masks, and impact of their precision on stochastic output, as well as impact of number of seeding points on tractography results.
:* At the same time, we are testing the module on a Max Plank dataset of Anna Rotarska-Jagiela by looking at the connections through the corpus callosum of the left auditory cortex to the right auditory cortex as defined by fMRI activation maps as well as running this data for the tractography comparison project.
:* We are discussing possibility of adding project specific functionality to module, such as third ROI to guide tractography, nonlinear registration button for merging fMRI and DTI data.

; B - Related Clinical Projects

;* Arcuate Fasciculus Extraction Project
[[Image:STArcuate.jpg|thumb|right|200px|<font size=1>Figure 7: The arcuate fasciculus including seed, midpoint and target ROI's.</font>]]
We have started the project of investigating Arcuate Fasciculus using Stochastic Tractography (Figure 7). This structure is especially important in both VCFS and schizophrenia, as it connects language related areas (Brocka and Wernicke's), and is involved in language processing quite disturbed in schizophrenia patients. It also can not be reliably traced using deterministic tractography.
:Project involves:
:* Whole brain segmentation, and automatic extraction of regions interconnected by Arcuate Fasciculus (Inferior frontal and Superior Temporal Gyri).
:* White matter segmentation, in order to prevent algorithm from traveling through the ventricles, where diffusivity is high.
:* Non-linear registration of labelmaps to the DTI space.
:* Seeding tracts. We have piloted it using 5000 seeds per voxel, however it is quite time consuming running it on even most powerful computers in the lab, so we have experimented with smaller number of seeds per voxel. We tested 1000 seeds, which gave virtually identical results.
:* Extracting path of interest, and calculating FA along the path for group comparison. Presentation of previous results for 7 schizophrenics and 12 control subjects, can be found here: [[Media:NAMIC_AHM_Arcuate.ppt|Progress Report Presentation]].
:* Results presentation and paper submission. Abstract was submitted and accepted for presentation at World Biological Psychiatry Symposium in Venice, Italy. Paper is in preparation.

;* Semantic Network Connectivity Project
We use combination of fMRI and DTI data to define and characterize functional and anatomical connectivity within the semantic processing network in schizophrenia.
:Project involves:
:* fMRI data analysis and identification of functional nodes involved in semantic processing in healthy controls and seubjects with schizophrenia
:* Analysis of functional connectivity (using FSL) between nodes of semantic network
:* Whole brain Voxel Based analysis of DTI data in same population
:* Use of stochastic tractography to identify connections between functional nodes
:* Correlational analysis involving anatomical and functional connectivity data.
:* Results presentation and paper submission. Data was presented at HBM conference in 2007, paper has been submitted to HBM.
[[Image:Anna.png|thumb|right|200px|<font size=1>Figure 8: Max Plank data showing tracts through the corpus connecting 2 cortical ROIs defined by fMRI activations.</font>]]

;* Study of Default Network
We have started collaboration with Department of Neurophysiology Max Planck Institute in Frankfurt (Anna Rotarska contact person). They have dataset containing DTI and resting state fMRI in schizophrenia, and want to use stochastic tractography to measure integrity of anatomical connections within the default network in schizophrenia. Their fMRI data has been co-registered with anatomical scans, and has been put it into DTI space. Figure 8 shows pilot data of the white matter connections through the corpus callosum between the left and the right auditory cortex ROIs as defined by fMRI data. This data will be also used to test robustness of our module, since data was collected on a different scanner (3T Philips).

;* Tractography Comparison Project
We are also working on a [http://www.na-mic.org/Wiki/index.php/SanteFe.Tractography.Conference tractography comparison project]dataset, where we apply stochastic tractography to phantom, as well as test dataset.

===Staffing Plan===

* Sylvain and Yogesh are the DBP resources charged with adapting the tools in the NA-MIC Kit to the DBP needs
* Doug Markant, our NAMIC RA has left the lab, and now Doug Terry, is a new NAMIC RA.
* Julien is our new NAMIC software engineer. He is responsible for improving the STM (Stochastic Tractography Module), and making sure software works with STM compliant datasets.
[[Link Progress| Development Progress]]
* Polina is the algorithm core contact
* Brad is the engineering core contact

===Schedule===

* '''10/2007''' - Optimization of Stochastic Tractography algorythm for 1.5T data.
* '''10/2007''' - Algorythm testing on Santa Fe data set and diffusion phantom.
* '''06/2008''' - Optimization of Stochastic Tractography algorythm for 3T data.
* '''11/2008''' - Slicer 3 module prototype using python.
* '''12/2008''' - Slicer 3 module official release
* '''12/2008''' - Documentation and packaging for dissemination.
* '''12/2008''' - Arcuate Fasciculus results.
* '''01/2009''' - Arcuate Fasciuclus first draft of the paper.
* '''05/2009''' - Distortion correction and nonlinear registration added to the module
* '''05/2009''' - Symposium on tractography, including stochastic methods at World Biological Psychiatry Symposium in Florence, Italy.
* '''05/2009''' - Presentation of Arcuate Fasciculus findings at World Biological Psychiatry Symposium in Florence, Italy.

===Team and Institute===
*PI: Marek Kubicki (kubicki at bwh.harvard.edu)
*DBP2 Investigators: Sylvain Bouix, Yogesh Rathi, Julien de Siebenthal
*NA-MIC Engineering Contact: Brad Davis, Kitware
*NA-MIC Algorithms Contact: Polina Gollard, MIT

===Publications===

''In print''

*[http://www.na-mic.org/pages/Special:Publications?text=Kubicki+AND+Westin+AND+DTI&submit=Search&words=all&title=checked&keywords=checked&authors=checked&abstract=checked&sponsors=checked&searchbytag=checked| NA-MIC Publications Database - Clinical Applications]

*[http://www.na-mic.org/pages/Special:Publications?text=Ngo+AND+Golland&submit=Search&words=all&title=checked&keywords=checked&authors=checked&abstract=checked&sponsors=checked&searchbytag=checked| NA-MIC Publications Database - Algorithms Development]

[[Category: Schizophrenia]] [[Category: Diffusion MRI]] [[Category: Segmentation]]

DBP2:Harvard:Brain Segmentation Roadmap

2009-06-03T18:27:32Z

Dougt: /* Module */

Back to [[NA-MIC_Internal_Collaborations|NA-MIC Collaborations]], [[DBP2:Harvard|Harvard DBP 2]]
__NOTOC__
=Stochastic Tractography for VCFS=
== Roadmap ==

The main goal of this project is to develop end-to-end application that would be used to characterize anatomical connectivity abnormalities in the brain of patients with velocardiofacial syndrome (VCFS), and to link this information with deficits in schizophrenia. This page describes the technology roadmap for stochastic tractography, using newly acquired 3T data, NAMIC tools and slicer 3.

== Algorithm ==

[[Image:IC_sto_new.png|thumb|right|200px|<font size=1> Figure 1: Comparison of deterministic and stochastic tractography algorithms</font>]]
; A-Description
* Most tractography methods estimate fibers by tracing the maximum direction of diffusion. A limitation of this approach is that, in practice, several factors introduce uncertainty in the tracking procedure, including, noise, splitting and crossing fibers, head motion and image artifacts. To address this uncertainty, stochastic tractography methods have been developed to quantify the uncertainty associated with estimated fibers (Bjornemo et al., 2002). Method uses a propagation model based on stochastics and regularization, which allows paths originating at one point to branch and return a probability distribution of possible paths. The method utilizes principles of a statistical Monte Carlo method called Sequential Importance Sampling and Resampling (SISR). Based on probability functions, using a sequential importance sampling technique ([http://lmi.bwh.harvard.edu/papers/pdfs/2002/bjornemoMICCAI02.pdf Bjornemo et al., 2002]), one can generate thousands of fibers starting in the same point by sequentially drawing random step directions. This gives a very rich model of the fiber distribution, as contrasted with single fibers produced by conventional tractography methods. Moreover, from a large number of sampled paths, probability maps can be generated, providing better estimates of connectivity between several anatomical locations. A comparison of the algorithms can be seen here. (Figure 1)

[[Image:StochasticPic.PNG|thumb|right|200px|<font size=1>Figure 2: Stochastic tractography of uncinate fasciculis on anatomical data (left) and cingulum bungle on fMRI scan (right)</font>]]
; B-Possible Applications
* Since diffusion direction uncertainty within the gray matter is quite significant; principal diffusion direction approaches usually do not work for tracking between two gray matter regions. Thus if one requires finding connections between a priori selected anatomical gray matter regions, defined either by anatomical segmentations (in case of using structural ROI data), or functional activations (in case of megring DTI with fMRI), stochastic tractography seems to be the method of choice. Here is an example of this application to anatomical data (Figure 2, left image) and to fMRI data (Figure 2, right image).

* Stochastic Tractography is also comparable, if not better, in defining large white matter fiber bundles, especially those traveling through white matter regions characterized by increased diffusion uncertainty (fiber crossings). Example of such application to internal capsule. (Figure 3)

[[Image:IC-comp-new.png|thumb|right|200px|<font size=1>Figure 3: Streamline vs. stochastic tractography of the Internal Capsule</font>]]
; C-References

* [http://lmi.bwh.harvard.edu/papers/pdfs/2002/bjornemoMICCAI02.pdf Björnemo M, Brun A, Kikinis R, Westin CF. Regularized stochastic white matter tractography using diffusion tensor MRI. In Fifth International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI'02). Tokyo, Japan, 2002;435-442.]
* [http://lmi.bwh.harvard.edu/papers/pdfs/2006/frimanTMI06.pdf Friman, O., Farneback, G., Westin CF. A Bayesian Approach for Stochastic White Matter Tractography. IEEE Transactions on Medical Imaging, Vol 25, No. 8, Aug. 2006]

[[Image:StochasticGUI1.PNG|thumb|right|200px|<font size=1> Figure 4: Python Stochastic Tractography GUI </font>]]

==Module==
Can be found in: MODULES > PYTHON MODULES > PYTHON STOCHASTIC TRACTOGRAPHY
;Functionality of Python Stochastic Tractography module in Slicer 3.0
* IO:
Module reads files (DWI and ROIs) in nhdr format.
* Smoothing:
One can smooth the DWI data (only Gausian smoothing is supported at this time). We recommend it if the data is noisy.
* Brain Mask:
The Brain mask defines the volume in which the tensor will be computed and the tracts evaluated.
*Diffusion Tensor:
This step calculates the tensor and can output anisotropy indices (FA/Mode/Trace)
*Tractography:
Parameters that need to be adjusted:
:1. The amount of tracts that will be seeded from each voxel (we recommend between 500 and 1000 tracts, depending on the workstation power- 1000 tracts per voxel seeded within the large ROI for high resolution DWI can take a long time to compute).
:2. Maximum tract length (in mm), this can eliminate long, unwanted tracts if the regions for which connection is measured are located close to each other
:3. Step Size: distance between each re-estimation of tensors, usually between 0.5 and 1 mm
:4. Stopping criteria. This can be used on the top of WM mask, to terminate tracts (in case they really want to travel through the CSF, for example).
*Connectivity Map:
This step creates output probability maps.
:1. binary: each voxel is counted only once if at least one fiber pass through it
:2. cumulative: tracts are summed by voxel independently
:3. weighted: tracts are summed by voxel depending on their length
*Length Based:
This step will output only either the shortest 1/3, middle 1/3, or longest 1/3 of the tracts.
*Vicinity
This step traces back n number of steps from tract endpoint to check if track crosses target ROI. If so, tract is included.
*Threshold
This step will reject tracts whose endpoints are lower than the threshold value).
*Spherical ROI vicinity
This will make the ROI a sphere based on the ROI’s center of gravity (with the sphere’s radius being the distance from the center to the ROI’s furthest point). This sphere can be inflated by raising the Vicinity level to the number of steps you’d like to increase the ROI’s size by.

Then, probability maps can be saved as ROIs, and either used directly, or thresholded (at certain probability, step claimed by few publications to remove noise) in slicer to mask and compute average FA, Mode, Trace for entire connection. Diffusion indices can be also weighted by the probability of connection for each voxel.

:* Module documentation can be found here:
:**[[Media:IJdata.tar.gz|Training Dataset]]
:**[[Media:Stochastic_June09.ppt|Training Presentation]]
:**[[Media:Helix.zip|Sample Helix Dataset]]

==Work Accomplished==
[[Image:helix_withsmoothing.png|thumb|right|200px|<font size=1>Figure 6: Stochastic tractography from a single ROI on helix phantom</font>]]
; A - Optimization and testing of stochastic tractography algorithm :
:* Original methodological paper, as well as our first attempts to use the algorithm (CC+ and matlab scripts) have been done on old "NAMIC" 1.5T LSDI data ([https://portal.nbirn.net:443/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=gotoDirectory&CPq_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FHarvard%2Farchive_morph&up=CPq&JavaScript=enabled Structural MRI] and [https://portal.nbirn.net:443/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=gotoDirectory&up=7li&7li_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FHarvard%2Farchive_diffusion&JavaScript=enabled DTI data]).
:* Tri worked hard on making sure algorithm works on new high resolution 3T data (available here: [https://portal.nbirn.net/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=moveUpDir&gv7_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FPNL%2F3T_strct_dti_fmri%2Fcase01017&up=gv7&JavaScript=enabled|PNL 3T Data]).
:* Tests have been done also on the spiral diffusion phantom, to make sure diffusion directions and scanner coordinates are handled properly by the algorithm (Figure 6).

; B - Clinical Applications

:* Algorithm was used to trace and analyze anterior limb of the internal capsule on 1.5T data. It generated reacher representation of frontal fiber projections, it also turned out to be more sensitive to group differences in white matter integrity that conventional deterministic tractography (see Figure 6).
:* Algorithm was also used to trace smaller white matter fiber tracts, such as Cingulum, Fornix, Uncinate Fasciculus, Arcuate Fasciculus on 3T "Santa Fe" dataset (http://www.na-mic.org/Wiki/index.php/SanteFe.Tractography.Conference)

; C - References

:* [http://www.na-mic.org/Wiki/index.php/Image:IC_posternew.png Shenton, M.E., Ngo, T., Rosenberger, G., Westin, C.F., Levitt, J.J., McCarley, R.W., Kubicki, M. Study of Thalamo-Cortical White Matter Fiber Tract Projections in Schizophrenia Using Diffusion Stochastic Tractography. Poster presented at the 46th Meeting of the American College of Neuropsychopharmacology, Boca Raton, FL, December 2007.]
:* [http://www.na-mic.org/Wiki/index.php/Media:dougt_poster.pdf Terry DP, Rausch AC, Alvarado JL, Melonakos ED, Markant D, Westin CF, Kikinis R, von Siebenthal J, Shenton ME, Kubicki M. White Matter Properties of Emotion Related Connections in Schizophrenia. Poster presented at the 2009 Mysell Poster Day, Dept. of Psychiatry, Harvard Medical School, April 2009]
:* [http://www.na-mic.org/Wiki/index.php/Media:Jorge_poster.pdf Alvarado JL, Terry DP, Markant D, Ngo T, Kikinis R, Westin CF, McCarley RW, Shenton ME, Kubicki M. Study of Language-Related White Matter Tract Connections in Schizophrenia using Diffusion Stochastic Tractography. Poster presented at the 2009 Mysell Poster Day, Dept. of Psychiatry, Harvard Medical School, April 2009]
:* Melonakos ED, Shenton ME, Markant D, Alvarado J, Westin CF, Kubicki M. White Matter Properties of Orbitofrontal Connections in Schizophrenia. Poster being presented at the 64th Meeting of the Society of Biological Psychiatry. Vancouver, BC. May 2009.
:* Kubicki, M. Khan, U., Bobrow, L., O'Donnell, L. Pieper, S. Westin, CF., Shenton, ME. New Methods for Assessing Whole Brain DTI Abnormalities in Schizophrenia. Presentation given at the International Congress of World Psychiatric Association. Florence, Italy. April 2009.
:* Kubicki, M., Markant, D., Ngo, T., Westin, CF., McCarley, RW., Shenton, ME. Study of Language Related White Matter Fiber Tract Projections in Schizophrenia Using Diffusion Stochastic Tractography. Presentation given at the International Congress of World Psychiatric Association. Florence, Italy. April 2009.

== Work in Progress ==

; A - Optimization and Testing of stochastic tractography module :
:* Julien is testing now separate components of work flow, including the masks, and impact of their precision on stochastic output, as well as impact of number of seeding points on tractography results.
:* At the same time, we are testing the module on a Max Plank dataset of Anna Rotarska-Jagiela by looking at the connections through the corpus callosum of the left auditory cortex to the right auditory cortex as defined by fMRI activation maps as well as running this data for the tractography comparison project.
:* We are discussing possibility of adding project specific functionality to module, such as third ROI to guide tractography, nonlinear registration button for merging fMRI and DTI data.

; B - Related Clinical Projects

;* Arcuate Fasciculus Extraction Project
[[Image:STArcuate.jpg|thumb|right|200px|<font size=1>Figure 7: The arcuate fasciculus including seed, midpoint and target ROI's.</font>]]
We have started the project of investigating Arcuate Fasciculus using Stochastic Tractography (Figure 7). This structure is especially important in both VCFS and schizophrenia, as it connects language related areas (Brocka and Wernicke's), and is involved in language processing quite disturbed in schizophrenia patients. It also can not be reliably traced using deterministic tractography.
:Project involves:
:* Whole brain segmentation, and automatic extraction of regions interconnected by Arcuate Fasciculus (Inferior frontal and Superior Temporal Gyri).
:* White matter segmentation, in order to prevent algorithm from traveling through the ventricles, where diffusivity is high.
:* Non-linear registration of labelmaps to the DTI space.
:* Seeding tracts. We have piloted it using 5000 seeds per voxel, however it is quite time consuming running it on even most powerful computers in the lab, so we have experimented with smaller number of seeds per voxel. We tested 1000 seeds, which gave virtually identical results.
:* Extracting path of interest, and calculating FA along the path for group comparison. Presentation of previous results for 7 schizophrenics and 12 control subjects, can be found here: [[Media:NAMIC_AHM_Arcuate.ppt|Progress Report Presentation]].
:* Results presentation and paper submission. Abstract was submitted and accepted for presentation at World Biological Psychiatry Symposium in Venice, Italy. Paper is in preparation.

;* Semantic Network Connectivity Project
We use combination of fMRI and DTI data to define and characterize functional and anatomical connectivity within the semantic processing network in schizophrenia.
:Project involves:
:* fMRI data analysis and identification of functional nodes involved in semantic processing in healthy controls and seubjects with schizophrenia
:* Analysis of functional connectivity (using FSL) between nodes of semantic network
:* Whole brain Voxel Based analysis of DTI data in same population
:* Use of stochastic tractography to identify connections between functional nodes
:* Correlational analysis involving anatomical and functional connectivity data.
:* Results presentation and paper submission. Data was presented at HBM conference in 2007, paper has been submitted to HBM.
[[Image:Anna.png|thumb|right|200px|<font size=1>Figure 8: Max Plank data showing tracts through the corpus connecting 2 cortical ROIs defined by fMRI activations.</font>]]

;* Study of Default Network
We have started collaboration with Department of Neurophysiology Max Planck Institute in Frankfurt (Anna Rotarska contact person). They have dataset containing DTI and resting state fMRI in schizophrenia, and want to use stochastic tractography to measure integrity of anatomical connections within the default network in schizophrenia. Their fMRI data has been co-registered with anatomical scans, and has been put it into DTI space. Figure 8 shows pilot data of the white matter connections through the corpus callosum between the left and the right auditory cortex ROIs as defined by fMRI data. This data will be also used to test robustness of our module, since data was collected on a different scanner (3T Philips).

;* Tractography Comparison Project
We are also working on a [http://www.na-mic.org/Wiki/index.php/SanteFe.Tractography.Conference tractography comparison project]dataset, where we apply stochastic tractography to phantom, as well as test dataset.

===Staffing Plan===

* Sylvain and Yogesh are the DBP resources charged with adapting the tools in the NA-MIC Kit to the DBP needs
* Doug Markant, our NAMIC RA has left the lab, and now Doug Terry, is a new NAMIC RA.
* Julien is our new NAMIC software engineer. He is responsible for improving the STM (Stochastic Tractography Module), and making sure software works with STM compliant datasets.
[[Link Progress| Development Progress]]
* Polina is the algorithm core contact
* Brad is the engineering core contact

===Schedule===

* '''10/2007''' - Optimization of Stochastic Tractography algorythm for 1.5T data.
* '''10/2007''' - Algorythm testing on Santa Fe data set and diffusion phantom.
* '''06/2008''' - Optimization of Stochastic Tractography algorythm for 3T data.
* '''11/2008''' - Slicer 3 module prototype using python.
* '''12/2008''' - Slicer 3 module official release
* '''12/2008''' - Documentation and packaging for dissemination.
* '''12/2008''' - Arcuate Fasciculus results.
* '''01/2009''' - Arcuate Fasciuclus first draft of the paper.
* '''05/2009''' - Distortion correction and nonlinear registration added to the module
* '''05/2009''' - Symposium on tractography, including stochastic methods at World Biological Psychiatry Symposium in Florence, Italy.
* '''05/2009''' - Presentation of Arcuate Fasciculus findings at World Biological Psychiatry Symposium in Florence, Italy.

===Team and Institute===
*PI: Marek Kubicki (kubicki at bwh.harvard.edu)
*DBP2 Investigators: Sylvain Bouix, Yogesh Rathi, Julien de Siebenthal
*NA-MIC Engineering Contact: Brad Davis, Kitware
*NA-MIC Algorithms Contact: Polina Gollard, MIT

===Publications===

''In print''

*[http://www.na-mic.org/pages/Special:Publications?text=Kubicki+AND+Westin+AND+DTI&submit=Search&words=all&title=checked&keywords=checked&authors=checked&abstract=checked&sponsors=checked&searchbytag=checked| NA-MIC Publications Database - Clinical Applications]

*[http://www.na-mic.org/pages/Special:Publications?text=Ngo+AND+Golland&submit=Search&words=all&title=checked&keywords=checked&authors=checked&abstract=checked&sponsors=checked&searchbytag=checked| NA-MIC Publications Database - Algorithms Development]

[[Category: Schizophrenia]] [[Category: Diffusion MRI]] [[Category: Segmentation]]

DBP2:Harvard:Brain Segmentation Roadmap

2009-06-03T18:16:07Z

Dougt:

Back to [[NA-MIC_Internal_Collaborations|NA-MIC Collaborations]], [[DBP2:Harvard|Harvard DBP 2]]
__NOTOC__
=Stochastic Tractography for VCFS=
== Roadmap ==

The main goal of this project is to develop end-to-end application that would be used to characterize anatomical connectivity abnormalities in the brain of patients with velocardiofacial syndrome (VCFS), and to link this information with deficits in schizophrenia. This page describes the technology roadmap for stochastic tractography, using newly acquired 3T data, NAMIC tools and slicer 3.

== Algorithm ==

[[Image:IC_sto_new.png|thumb|right|200px|<font size=1> Figure 1: Comparison of deterministic and stochastic tractography algorithms</font>]]
; A-Description
* Most tractography methods estimate fibers by tracing the maximum direction of diffusion. A limitation of this approach is that, in practice, several factors introduce uncertainty in the tracking procedure, including, noise, splitting and crossing fibers, head motion and image artifacts. To address this uncertainty, stochastic tractography methods have been developed to quantify the uncertainty associated with estimated fibers (Bjornemo et al., 2002). Method uses a propagation model based on stochastics and regularization, which allows paths originating at one point to branch and return a probability distribution of possible paths. The method utilizes principles of a statistical Monte Carlo method called Sequential Importance Sampling and Resampling (SISR). Based on probability functions, using a sequential importance sampling technique ([http://lmi.bwh.harvard.edu/papers/pdfs/2002/bjornemoMICCAI02.pdf Bjornemo et al., 2002]), one can generate thousands of fibers starting in the same point by sequentially drawing random step directions. This gives a very rich model of the fiber distribution, as contrasted with single fibers produced by conventional tractography methods. Moreover, from a large number of sampled paths, probability maps can be generated, providing better estimates of connectivity between several anatomical locations. A comparison of the algorithms can be seen here. (Figure 1)

[[Image:StochasticPic.PNG|thumb|right|200px|<font size=1>Figure 2: Stochastic tractography of uncinate fasciculis on anatomical data (left) and cingulum bungle on fMRI scan (right)</font>]]
; B-Possible Applications
* Since diffusion direction uncertainty within the gray matter is quite significant; principal diffusion direction approaches usually do not work for tracking between two gray matter regions. Thus if one requires finding connections between a priori selected anatomical gray matter regions, defined either by anatomical segmentations (in case of using structural ROI data), or functional activations (in case of megring DTI with fMRI), stochastic tractography seems to be the method of choice. Here is an example of this application to anatomical data (Figure 2, left image) and to fMRI data (Figure 2, right image).

* Stochastic Tractography is also comparable, if not better, in defining large white matter fiber bundles, especially those traveling through white matter regions characterized by increased diffusion uncertainty (fiber crossings). Example of such application to internal capsule. (Figure 3)

[[Image:IC-comp-new.png|thumb|right|200px|<font size=1>Figure 3: Streamline vs. stochastic tractography of the Internal Capsule</font>]]
; C-References

* [http://lmi.bwh.harvard.edu/papers/pdfs/2002/bjornemoMICCAI02.pdf Björnemo M, Brun A, Kikinis R, Westin CF. Regularized stochastic white matter tractography using diffusion tensor MRI. In Fifth International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI'02). Tokyo, Japan, 2002;435-442.]
* [http://lmi.bwh.harvard.edu/papers/pdfs/2006/frimanTMI06.pdf Friman, O., Farneback, G., Westin CF. A Bayesian Approach for Stochastic White Matter Tractography. IEEE Transactions on Medical Imaging, Vol 25, No. 8, Aug. 2006]

[[Image:StochasticGUI1.PNG|thumb|right|200px|<font size=1> Figure 4: Python Stochastic Tractography GUI </font>]]

==Module==
Can be found in: MODULES > PYTHON MODULES > PYTHON STOCHASTIC TRACTOGRAPHY
;Functionality of Python Stochastic Tractography module in Slicer 3.0
* IO:
Module reads files (DWI and ROIs) in nhdr format.
* Smoothing:
One can smooth the DWI data (only Gausian smoothing is supported at this time). We recommend it if the data is noisy.
* Brain Mask:
The Brain mask defines the volume in which the tensor will be computed and the tracts evaluated.
*Diffusion Tensor:
This step calculates the tensor and can output anisotropy indices (FA/Mode/Trace)
*Tractography:
Parameters that need to be adjusted:
:1. The amount of tracts that will be seeded from each voxel (we recommend between 500 and 1000 tracts, depending on the workstation power- 1000 tracts per voxel seeded within the large ROI for high resolution DWI can take a long time to compute).
:2. Maximum tract length (in mm), this can eliminate long, unwanted tracts if the regions for which connection is measured are located close to each other
:3. Step Size: distance between each re-estimation of tensors, usually between 0.5 and 1 mm
:4. Stopping criteria. This can be used on the top of WM mask, to terminate tracts (in case they really want to travel through the CSF, for example).
*Connectivity Map:
This step creates output probability maps.
:1. binary: each voxel is counted only once if at least one fiber pass through it
:2. cumulative: tracts are summed by voxel independently
:3. weighted: tracts are summed by voxel depending on their length

Then, probability maps can be saved as ROIs, and either used directly, or thresholded (at certain probability, step claimed by few publications to remove noise) in slicer to mask and compute average FA, Mode, Trace for entire connection. Diffusion indices can be also weighted by the probability of connection for each voxel.

:* Module documentation can be found here:
:**[[Media:IJdata.tar.gz|Training Dataset]]
:**[[Media:Stochastic_June09.ppt|Training Presentation]]
:**[[Media:Helix.zip|Sample Helix Dataset]]

==Work Accomplished==
[[Image:helix_withsmoothing.png|thumb|right|200px|<font size=1>Figure 6: Stochastic tractography from a single ROI on helix phantom</font>]]
; A - Optimization and testing of stochastic tractography algorithm :
:* Original methodological paper, as well as our first attempts to use the algorithm (CC+ and matlab scripts) have been done on old "NAMIC" 1.5T LSDI data ([https://portal.nbirn.net:443/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=gotoDirectory&CPq_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FHarvard%2Farchive_morph&up=CPq&JavaScript=enabled Structural MRI] and [https://portal.nbirn.net:443/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=gotoDirectory&up=7li&7li_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FHarvard%2Farchive_diffusion&JavaScript=enabled DTI data]).
:* Tri worked hard on making sure algorithm works on new high resolution 3T data (available here: [https://portal.nbirn.net/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=moveUpDir&gv7_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FPNL%2F3T_strct_dti_fmri%2Fcase01017&up=gv7&JavaScript=enabled|PNL 3T Data]).
:* Tests have been done also on the spiral diffusion phantom, to make sure diffusion directions and scanner coordinates are handled properly by the algorithm (Figure 6).

; B - Clinical Applications

:* Algorithm was used to trace and analyze anterior limb of the internal capsule on 1.5T data. It generated reacher representation of frontal fiber projections, it also turned out to be more sensitive to group differences in white matter integrity that conventional deterministic tractography (see Figure 6).
:* Algorithm was also used to trace smaller white matter fiber tracts, such as Cingulum, Fornix, Uncinate Fasciculus, Arcuate Fasciculus on 3T "Santa Fe" dataset (http://www.na-mic.org/Wiki/index.php/SanteFe.Tractography.Conference)

; C - References

:* [http://www.na-mic.org/Wiki/index.php/Image:IC_posternew.png Shenton, M.E., Ngo, T., Rosenberger, G., Westin, C.F., Levitt, J.J., McCarley, R.W., Kubicki, M. Study of Thalamo-Cortical White Matter Fiber Tract Projections in Schizophrenia Using Diffusion Stochastic Tractography. Poster presented at the 46th Meeting of the American College of Neuropsychopharmacology, Boca Raton, FL, December 2007.]
:* [http://www.na-mic.org/Wiki/index.php/Media:dougt_poster.pdf Terry DP, Rausch AC, Alvarado JL, Melonakos ED, Markant D, Westin CF, Kikinis R, von Siebenthal J, Shenton ME, Kubicki M. White Matter Properties of Emotion Related Connections in Schizophrenia. Poster presented at the 2009 Mysell Poster Day, Dept. of Psychiatry, Harvard Medical School, April 2009]
:* [http://www.na-mic.org/Wiki/index.php/Media:Jorge_poster.pdf Alvarado JL, Terry DP, Markant D, Ngo T, Kikinis R, Westin CF, McCarley RW, Shenton ME, Kubicki M. Study of Language-Related White Matter Tract Connections in Schizophrenia using Diffusion Stochastic Tractography. Poster presented at the 2009 Mysell Poster Day, Dept. of Psychiatry, Harvard Medical School, April 2009]
:* Melonakos ED, Shenton ME, Markant D, Alvarado J, Westin CF, Kubicki M. White Matter Properties of Orbitofrontal Connections in Schizophrenia. Poster being presented at the 64th Meeting of the Society of Biological Psychiatry. Vancouver, BC. May 2009.
:* Kubicki, M. Khan, U., Bobrow, L., O'Donnell, L. Pieper, S. Westin, CF., Shenton, ME. New Methods for Assessing Whole Brain DTI Abnormalities in Schizophrenia. Presentation given at the International Congress of World Psychiatric Association. Florence, Italy. April 2009.
:* Kubicki, M., Markant, D., Ngo, T., Westin, CF., McCarley, RW., Shenton, ME. Study of Language Related White Matter Fiber Tract Projections in Schizophrenia Using Diffusion Stochastic Tractography. Presentation given at the International Congress of World Psychiatric Association. Florence, Italy. April 2009.

== Work in Progress ==

; A - Optimization and Testing of stochastic tractography module :
:* Julien is testing now separate components of work flow, including the masks, and impact of their precision on stochastic output, as well as impact of number of seeding points on tractography results.
:* At the same time, we are testing the module on a Max Plank dataset of Anna Rotarska-Jagiela by looking at the connections through the corpus callosum of the left auditory cortex to the right auditory cortex as defined by fMRI activation maps as well as running this data for the tractography comparison project.
:* We are discussing possibility of adding project specific functionality to module, such as third ROI to guide tractography, nonlinear registration button for merging fMRI and DTI data.

; B - Related Clinical Projects

;* Arcuate Fasciculus Extraction Project
[[Image:STArcuate.jpg|thumb|right|200px|<font size=1>Figure 7: The arcuate fasciculus including seed, midpoint and target ROI's.</font>]]
We have started the project of investigating Arcuate Fasciculus using Stochastic Tractography (Figure 7). This structure is especially important in both VCFS and schizophrenia, as it connects language related areas (Brocka and Wernicke's), and is involved in language processing quite disturbed in schizophrenia patients. It also can not be reliably traced using deterministic tractography.
:Project involves:
:* Whole brain segmentation, and automatic extraction of regions interconnected by Arcuate Fasciculus (Inferior frontal and Superior Temporal Gyri).
:* White matter segmentation, in order to prevent algorithm from traveling through the ventricles, where diffusivity is high.
:* Non-linear registration of labelmaps to the DTI space.
:* Seeding tracts. We have piloted it using 5000 seeds per voxel, however it is quite time consuming running it on even most powerful computers in the lab, so we have experimented with smaller number of seeds per voxel. We tested 1000 seeds, which gave virtually identical results.
:* Extracting path of interest, and calculating FA along the path for group comparison. Presentation of previous results for 7 schizophrenics and 12 control subjects, can be found here: [[Media:NAMIC_AHM_Arcuate.ppt|Progress Report Presentation]].
:* Results presentation and paper submission. Abstract was submitted and accepted for presentation at World Biological Psychiatry Symposium in Venice, Italy. Paper is in preparation.

;* Semantic Network Connectivity Project
We use combination of fMRI and DTI data to define and characterize functional and anatomical connectivity within the semantic processing network in schizophrenia.
:Project involves:
:* fMRI data analysis and identification of functional nodes involved in semantic processing in healthy controls and seubjects with schizophrenia
:* Analysis of functional connectivity (using FSL) between nodes of semantic network
:* Whole brain Voxel Based analysis of DTI data in same population
:* Use of stochastic tractography to identify connections between functional nodes
:* Correlational analysis involving anatomical and functional connectivity data.
:* Results presentation and paper submission. Data was presented at HBM conference in 2007, paper has been submitted to HBM.
[[Image:Anna.png|thumb|right|200px|<font size=1>Figure 8: Max Plank data showing tracts through the corpus connecting 2 cortical ROIs defined by fMRI activations.</font>]]

;* Study of Default Network
We have started collaboration with Department of Neurophysiology Max Planck Institute in Frankfurt (Anna Rotarska contact person). They have dataset containing DTI and resting state fMRI in schizophrenia, and want to use stochastic tractography to measure integrity of anatomical connections within the default network in schizophrenia. Their fMRI data has been co-registered with anatomical scans, and has been put it into DTI space. Figure 8 shows pilot data of the white matter connections through the corpus callosum between the left and the right auditory cortex ROIs as defined by fMRI data. This data will be also used to test robustness of our module, since data was collected on a different scanner (3T Philips).

;* Tractography Comparison Project
We are also working on a [http://www.na-mic.org/Wiki/index.php/SanteFe.Tractography.Conference tractography comparison project]dataset, where we apply stochastic tractography to phantom, as well as test dataset.

===Staffing Plan===

* Sylvain and Yogesh are the DBP resources charged with adapting the tools in the NA-MIC Kit to the DBP needs
* Doug Markant, our NAMIC RA has left the lab, and now Doug Terry, is a new NAMIC RA.
* Julien is our new NAMIC software engineer. He is responsible for improving the STM (Stochastic Tractography Module), and making sure software works with STM compliant datasets.
[[Link Progress| Development Progress]]
* Polina is the algorithm core contact
* Brad is the engineering core contact

===Schedule===

* '''10/2007''' - Optimization of Stochastic Tractography algorythm for 1.5T data.
* '''10/2007''' - Algorythm testing on Santa Fe data set and diffusion phantom.
* '''06/2008''' - Optimization of Stochastic Tractography algorythm for 3T data.
* '''11/2008''' - Slicer 3 module prototype using python.
* '''12/2008''' - Slicer 3 module official release
* '''12/2008''' - Documentation and packaging for dissemination.
* '''12/2008''' - Arcuate Fasciculus results.
* '''01/2009''' - Arcuate Fasciuclus first draft of the paper.
* '''05/2009''' - Distortion correction and nonlinear registration added to the module
* '''05/2009''' - Symposium on tractography, including stochastic methods at World Biological Psychiatry Symposium in Florence, Italy.
* '''05/2009''' - Presentation of Arcuate Fasciculus findings at World Biological Psychiatry Symposium in Florence, Italy.

===Team and Institute===
*PI: Marek Kubicki (kubicki at bwh.harvard.edu)
*DBP2 Investigators: Sylvain Bouix, Yogesh Rathi, Julien de Siebenthal
*NA-MIC Engineering Contact: Brad Davis, Kitware
*NA-MIC Algorithms Contact: Polina Gollard, MIT

===Publications===

''In print''

*[http://www.na-mic.org/pages/Special:Publications?text=Kubicki+AND+Westin+AND+DTI&submit=Search&words=all&title=checked&keywords=checked&authors=checked&abstract=checked&sponsors=checked&searchbytag=checked| NA-MIC Publications Database - Clinical Applications]

*[http://www.na-mic.org/pages/Special:Publications?text=Ngo+AND+Golland&submit=Search&words=all&title=checked&keywords=checked&authors=checked&abstract=checked&sponsors=checked&searchbytag=checked| NA-MIC Publications Database - Algorithms Development]

[[Category: Schizophrenia]] [[Category: Diffusion MRI]] [[Category: Segmentation]]

DBP2:Harvard:Brain Segmentation Roadmap

2009-06-03T18:15:41Z

Dougt:

Back to [[NA-MIC_Internal_Collaborations|NA-MIC Collaborations]], [[DBP2:Harvard|Harvard DBP 2]]
__NOTOC__
=Stochastic Tractography for VCFS=
== Roadmap ==

The main goal of this project is to develop end-to-end application that would be used to characterize anatomical connectivity abnormalities in the brain of patients with velocardiofacial syndrome (VCFS), and to link this information with deficits in schizophrenia. This page describes the technology roadmap for stochastic tractography, using newly acquired 3T data, NAMIC tools and slicer 3.

== Algorithm ==

[[Image:IC_sto_new.png|thumb|right|200px|<font size=1> Figure 1: Comparison of deterministic and stochastic tractography algorithms</font>]]
; A-Description
* Most tractography methods estimate fibers by tracing the maximum direction of diffusion. A limitation of this approach is that, in practice, several factors introduce uncertainty in the tracking procedure, including, noise, splitting and crossing fibers, head motion and image artifacts. To address this uncertainty, stochastic tractography methods have been developed to quantify the uncertainty associated with estimated fibers (Bjornemo et al., 2002). Method uses a propagation model based on stochastics and regularization, which allows paths originating at one point to branch and return a probability distribution of possible paths. The method utilizes principles of a statistical Monte Carlo method called Sequential Importance Sampling and Resampling (SISR). Based on probability functions, using a sequential importance sampling technique ([http://lmi.bwh.harvard.edu/papers/pdfs/2002/bjornemoMICCAI02.pdf Bjornemo et al., 2002]), one can generate thousands of fibers starting in the same point by sequentially drawing random step directions. This gives a very rich model of the fiber distribution, as contrasted with single fibers produced by conventional tractography methods. Moreover, from a large number of sampled paths, probability maps can be generated, providing better estimates of connectivity between several anatomical locations. A comparison of the algorithms can be seen here. (Figure 1)

[[Image:StochasticPic.PNG|thumb|right|200px|<font size=1>Figure 2: Stochastic tractography of uncinate fasciculis on anatomical data (left) and cingulum bungle on fMRI scan (right)</font>]]
; B-Possible Applications
* Since diffusion direction uncertainty within the gray matter is quite significant; principal diffusion direction approaches usually do not work for tracking between two gray matter regions. Thus if one requires finding connections between a priori selected anatomical gray matter regions, defined either by anatomical segmentations (in case of using structural ROI data), or functional activations (in case of megring DTI with fMRI), stochastic tractography seems to be the method of choice. Here is an example of this application to anatomical data (Figure 2, left image) and to fMRI data (Figure 2, right image).

* Stochastic Tractography is also comparable, if not better, in defining large white matter fiber bundles, especially those traveling through white matter regions characterized by increased diffusion uncertainty (fiber crossings). Example of such application to internal capsule. (Figure 3)

[[Image:IC-comp-new.png|thumb|right|200px|<font size=1>Figure 3: Streamline vs. stochastic tractography of the Internal Capsule</font>]]
; C-References

* [http://lmi.bwh.harvard.edu/papers/pdfs/2002/bjornemoMICCAI02.pdf Björnemo M, Brun A, Kikinis R, Westin CF. Regularized stochastic white matter tractography using diffusion tensor MRI. In Fifth International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI'02). Tokyo, Japan, 2002;435-442.]
* [http://lmi.bwh.harvard.edu/papers/pdfs/2006/frimanTMI06.pdf Friman, O., Farneback, G., Westin CF. A Bayesian Approach for Stochastic White Matter Tractography. IEEE Transactions on Medical Imaging, Vol 25, No. 8, Aug. 2006]

[[Image:StochasticGUI1.png|thumb|right|200px|<font size=1> Figure 4: Python Stochastic Tractography GUI </font>]]

==Module==
Can be found in: MODULES > PYTHON MODULES > PYTHON STOCHASTIC TRACTOGRAPHY
;Functionality of Python Stochastic Tractography module in Slicer 3.0
* IO:
Module reads files (DWI and ROIs) in nhdr format.
* Smoothing:
One can smooth the DWI data (only Gausian smoothing is supported at this time). We recommend it if the data is noisy.
* Brain Mask:
The Brain mask defines the volume in which the tensor will be computed and the tracts evaluated.
*Diffusion Tensor:
This step calculates the tensor and can output anisotropy indices (FA/Mode/Trace)
*Tractography:
Parameters that need to be adjusted:
:1. The amount of tracts that will be seeded from each voxel (we recommend between 500 and 1000 tracts, depending on the workstation power- 1000 tracts per voxel seeded within the large ROI for high resolution DWI can take a long time to compute).
:2. Maximum tract length (in mm), this can eliminate long, unwanted tracts if the regions for which connection is measured are located close to each other
:3. Step Size: distance between each re-estimation of tensors, usually between 0.5 and 1 mm
:4. Stopping criteria. This can be used on the top of WM mask, to terminate tracts (in case they really want to travel through the CSF, for example).
*Connectivity Map:
This step creates output probability maps.
:1. binary: each voxel is counted only once if at least one fiber pass through it
:2. cumulative: tracts are summed by voxel independently
:3. weighted: tracts are summed by voxel depending on their length

Then, probability maps can be saved as ROIs, and either used directly, or thresholded (at certain probability, step claimed by few publications to remove noise) in slicer to mask and compute average FA, Mode, Trace for entire connection. Diffusion indices can be also weighted by the probability of connection for each voxel.

:* Module documentation can be found here:
:**[[Media:IJdata.tar.gz|Training Dataset]]
:**[[Media:Stochastic_June09.ppt|Training Presentation]]
:**[[Media:Helix.zip|Sample Helix Dataset]]

==Work Accomplished==
[[Image:helix_withsmoothing.png|thumb|right|200px|<font size=1>Figure 6: Stochastic tractography from a single ROI on helix phantom</font>]]
; A - Optimization and testing of stochastic tractography algorithm :
:* Original methodological paper, as well as our first attempts to use the algorithm (CC+ and matlab scripts) have been done on old "NAMIC" 1.5T LSDI data ([https://portal.nbirn.net:443/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=gotoDirectory&CPq_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FHarvard%2Farchive_morph&up=CPq&JavaScript=enabled Structural MRI] and [https://portal.nbirn.net:443/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=gotoDirectory&up=7li&7li_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FHarvard%2Farchive_diffusion&JavaScript=enabled DTI data]).
:* Tri worked hard on making sure algorithm works on new high resolution 3T data (available here: [https://portal.nbirn.net/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=moveUpDir&gv7_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FPNL%2F3T_strct_dti_fmri%2Fcase01017&up=gv7&JavaScript=enabled|PNL 3T Data]).
:* Tests have been done also on the spiral diffusion phantom, to make sure diffusion directions and scanner coordinates are handled properly by the algorithm (Figure 6).

; B - Clinical Applications

:* Algorithm was used to trace and analyze anterior limb of the internal capsule on 1.5T data. It generated reacher representation of frontal fiber projections, it also turned out to be more sensitive to group differences in white matter integrity that conventional deterministic tractography (see Figure 6).
:* Algorithm was also used to trace smaller white matter fiber tracts, such as Cingulum, Fornix, Uncinate Fasciculus, Arcuate Fasciculus on 3T "Santa Fe" dataset (http://www.na-mic.org/Wiki/index.php/SanteFe.Tractography.Conference)

; C - References

:* [http://www.na-mic.org/Wiki/index.php/Image:IC_posternew.png Shenton, M.E., Ngo, T., Rosenberger, G., Westin, C.F., Levitt, J.J., McCarley, R.W., Kubicki, M. Study of Thalamo-Cortical White Matter Fiber Tract Projections in Schizophrenia Using Diffusion Stochastic Tractography. Poster presented at the 46th Meeting of the American College of Neuropsychopharmacology, Boca Raton, FL, December 2007.]
:* [http://www.na-mic.org/Wiki/index.php/Media:dougt_poster.pdf Terry DP, Rausch AC, Alvarado JL, Melonakos ED, Markant D, Westin CF, Kikinis R, von Siebenthal J, Shenton ME, Kubicki M. White Matter Properties of Emotion Related Connections in Schizophrenia. Poster presented at the 2009 Mysell Poster Day, Dept. of Psychiatry, Harvard Medical School, April 2009]
:* [http://www.na-mic.org/Wiki/index.php/Media:Jorge_poster.pdf Alvarado JL, Terry DP, Markant D, Ngo T, Kikinis R, Westin CF, McCarley RW, Shenton ME, Kubicki M. Study of Language-Related White Matter Tract Connections in Schizophrenia using Diffusion Stochastic Tractography. Poster presented at the 2009 Mysell Poster Day, Dept. of Psychiatry, Harvard Medical School, April 2009]
:* Melonakos ED, Shenton ME, Markant D, Alvarado J, Westin CF, Kubicki M. White Matter Properties of Orbitofrontal Connections in Schizophrenia. Poster being presented at the 64th Meeting of the Society of Biological Psychiatry. Vancouver, BC. May 2009.
:* Kubicki, M. Khan, U., Bobrow, L., O'Donnell, L. Pieper, S. Westin, CF., Shenton, ME. New Methods for Assessing Whole Brain DTI Abnormalities in Schizophrenia. Presentation given at the International Congress of World Psychiatric Association. Florence, Italy. April 2009.
:* Kubicki, M., Markant, D., Ngo, T., Westin, CF., McCarley, RW., Shenton, ME. Study of Language Related White Matter Fiber Tract Projections in Schizophrenia Using Diffusion Stochastic Tractography. Presentation given at the International Congress of World Psychiatric Association. Florence, Italy. April 2009.

== Work in Progress ==

; A - Optimization and Testing of stochastic tractography module :
:* Julien is testing now separate components of work flow, including the masks, and impact of their precision on stochastic output, as well as impact of number of seeding points on tractography results.
:* At the same time, we are testing the module on a Max Plank dataset of Anna Rotarska-Jagiela by looking at the connections through the corpus callosum of the left auditory cortex to the right auditory cortex as defined by fMRI activation maps as well as running this data for the tractography comparison project.
:* We are discussing possibility of adding project specific functionality to module, such as third ROI to guide tractography, nonlinear registration button for merging fMRI and DTI data.

; B - Related Clinical Projects

;* Arcuate Fasciculus Extraction Project
[[Image:STArcuate.jpg|thumb|right|200px|<font size=1>Figure 7: The arcuate fasciculus including seed, midpoint and target ROI's.</font>]]
We have started the project of investigating Arcuate Fasciculus using Stochastic Tractography (Figure 7). This structure is especially important in both VCFS and schizophrenia, as it connects language related areas (Brocka and Wernicke's), and is involved in language processing quite disturbed in schizophrenia patients. It also can not be reliably traced using deterministic tractography.
:Project involves:
:* Whole brain segmentation, and automatic extraction of regions interconnected by Arcuate Fasciculus (Inferior frontal and Superior Temporal Gyri).
:* White matter segmentation, in order to prevent algorithm from traveling through the ventricles, where diffusivity is high.
:* Non-linear registration of labelmaps to the DTI space.
:* Seeding tracts. We have piloted it using 5000 seeds per voxel, however it is quite time consuming running it on even most powerful computers in the lab, so we have experimented with smaller number of seeds per voxel. We tested 1000 seeds, which gave virtually identical results.
:* Extracting path of interest, and calculating FA along the path for group comparison. Presentation of previous results for 7 schizophrenics and 12 control subjects, can be found here: [[Media:NAMIC_AHM_Arcuate.ppt|Progress Report Presentation]].
:* Results presentation and paper submission. Abstract was submitted and accepted for presentation at World Biological Psychiatry Symposium in Venice, Italy. Paper is in preparation.

;* Semantic Network Connectivity Project
We use combination of fMRI and DTI data to define and characterize functional and anatomical connectivity within the semantic processing network in schizophrenia.
:Project involves:
:* fMRI data analysis and identification of functional nodes involved in semantic processing in healthy controls and seubjects with schizophrenia
:* Analysis of functional connectivity (using FSL) between nodes of semantic network
:* Whole brain Voxel Based analysis of DTI data in same population
:* Use of stochastic tractography to identify connections between functional nodes
:* Correlational analysis involving anatomical and functional connectivity data.
:* Results presentation and paper submission. Data was presented at HBM conference in 2007, paper has been submitted to HBM.
[[Image:Anna.png|thumb|right|200px|<font size=1>Figure 8: Max Plank data showing tracts through the corpus connecting 2 cortical ROIs defined by fMRI activations.</font>]]

;* Study of Default Network
We have started collaboration with Department of Neurophysiology Max Planck Institute in Frankfurt (Anna Rotarska contact person). They have dataset containing DTI and resting state fMRI in schizophrenia, and want to use stochastic tractography to measure integrity of anatomical connections within the default network in schizophrenia. Their fMRI data has been co-registered with anatomical scans, and has been put it into DTI space. Figure 8 shows pilot data of the white matter connections through the corpus callosum between the left and the right auditory cortex ROIs as defined by fMRI data. This data will be also used to test robustness of our module, since data was collected on a different scanner (3T Philips).

;* Tractography Comparison Project
We are also working on a [http://www.na-mic.org/Wiki/index.php/SanteFe.Tractography.Conference tractography comparison project]dataset, where we apply stochastic tractography to phantom, as well as test dataset.

===Staffing Plan===

* Sylvain and Yogesh are the DBP resources charged with adapting the tools in the NA-MIC Kit to the DBP needs
* Doug Markant, our NAMIC RA has left the lab, and now Doug Terry, is a new NAMIC RA.
* Julien is our new NAMIC software engineer. He is responsible for improving the STM (Stochastic Tractography Module), and making sure software works with STM compliant datasets.
[[Link Progress| Development Progress]]
* Polina is the algorithm core contact
* Brad is the engineering core contact

===Schedule===

* '''10/2007''' - Optimization of Stochastic Tractography algorythm for 1.5T data.
* '''10/2007''' - Algorythm testing on Santa Fe data set and diffusion phantom.
* '''06/2008''' - Optimization of Stochastic Tractography algorythm for 3T data.
* '''11/2008''' - Slicer 3 module prototype using python.
* '''12/2008''' - Slicer 3 module official release
* '''12/2008''' - Documentation and packaging for dissemination.
* '''12/2008''' - Arcuate Fasciculus results.
* '''01/2009''' - Arcuate Fasciuclus first draft of the paper.
* '''05/2009''' - Distortion correction and nonlinear registration added to the module
* '''05/2009''' - Symposium on tractography, including stochastic methods at World Biological Psychiatry Symposium in Florence, Italy.
* '''05/2009''' - Presentation of Arcuate Fasciculus findings at World Biological Psychiatry Symposium in Florence, Italy.

===Team and Institute===
*PI: Marek Kubicki (kubicki at bwh.harvard.edu)
*DBP2 Investigators: Sylvain Bouix, Yogesh Rathi, Julien de Siebenthal
*NA-MIC Engineering Contact: Brad Davis, Kitware
*NA-MIC Algorithms Contact: Polina Gollard, MIT

===Publications===

''In print''

*[http://www.na-mic.org/pages/Special:Publications?text=Kubicki+AND+Westin+AND+DTI&submit=Search&words=all&title=checked&keywords=checked&authors=checked&abstract=checked&sponsors=checked&searchbytag=checked| NA-MIC Publications Database - Clinical Applications]

*[http://www.na-mic.org/pages/Special:Publications?text=Ngo+AND+Golland&submit=Search&words=all&title=checked&keywords=checked&authors=checked&abstract=checked&sponsors=checked&searchbytag=checked| NA-MIC Publications Database - Algorithms Development]

[[Category: Schizophrenia]] [[Category: Diffusion MRI]] [[Category: Segmentation]]

File:StochasticGUI1.PNG

2009-06-03T18:15:10Z

Dougt:

DBP2:Harvard:Brain Segmentation Roadmap

2009-06-03T18:11:40Z

Dougt:

Back to [[NA-MIC_Internal_Collaborations|NA-MIC Collaborations]], [[DBP2:Harvard|Harvard DBP 2]]
__NOTOC__
=Stochastic Tractography for VCFS=
== Roadmap ==

The main goal of this project is to develop end-to-end application that would be used to characterize anatomical connectivity abnormalities in the brain of patients with velocardiofacial syndrome (VCFS), and to link this information with deficits in schizophrenia. This page describes the technology roadmap for stochastic tractography, using newly acquired 3T data, NAMIC tools and slicer 3.

== Algorithm ==

[[Image:IC_sto_new.png|thumb|right|200px|<font size=1> Figure 1: Comparison of deterministic and stochastic tractography algorithms</font>]]
; A-Description
* Most tractography methods estimate fibers by tracing the maximum direction of diffusion. A limitation of this approach is that, in practice, several factors introduce uncertainty in the tracking procedure, including, noise, splitting and crossing fibers, head motion and image artifacts. To address this uncertainty, stochastic tractography methods have been developed to quantify the uncertainty associated with estimated fibers (Bjornemo et al., 2002). Method uses a propagation model based on stochastics and regularization, which allows paths originating at one point to branch and return a probability distribution of possible paths. The method utilizes principles of a statistical Monte Carlo method called Sequential Importance Sampling and Resampling (SISR). Based on probability functions, using a sequential importance sampling technique ([http://lmi.bwh.harvard.edu/papers/pdfs/2002/bjornemoMICCAI02.pdf Bjornemo et al., 2002]), one can generate thousands of fibers starting in the same point by sequentially drawing random step directions. This gives a very rich model of the fiber distribution, as contrasted with single fibers produced by conventional tractography methods. Moreover, from a large number of sampled paths, probability maps can be generated, providing better estimates of connectivity between several anatomical locations. A comparison of the algorithms can be seen here. (Figure 1)

[[Image:StochasticPic.PNG|thumb|right|200px|<font size=1>Figure 2: Stochastic tractography of uncinate fasciculis on anatomical data (left) and cingulum bungle on fMRI scan (right)</font>]]
; B-Possible Applications
* Since diffusion direction uncertainty within the gray matter is quite significant; principal diffusion direction approaches usually do not work for tracking between two gray matter regions. Thus if one requires finding connections between a priori selected anatomical gray matter regions, defined either by anatomical segmentations (in case of using structural ROI data), or functional activations (in case of megring DTI with fMRI), stochastic tractography seems to be the method of choice. Here is an example of this application to anatomical data (Figure 2, left image) and to fMRI data (Figure 2, right image).

* Stochastic Tractography is also comparable, if not better, in defining large white matter fiber bundles, especially those traveling through white matter regions characterized by increased diffusion uncertainty (fiber crossings). Example of such application to internal capsule. (Figure 3)

[[Image:IC-comp-new.png|thumb|right|200px|<font size=1>Figure 3: Streamline vs. stochastic tractography of the Internal Capsule</font>]]
; C-References

* [http://lmi.bwh.harvard.edu/papers/pdfs/2002/bjornemoMICCAI02.pdf Björnemo M, Brun A, Kikinis R, Westin CF. Regularized stochastic white matter tractography using diffusion tensor MRI. In Fifth International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI'02). Tokyo, Japan, 2002;435-442.]
* [http://lmi.bwh.harvard.edu/papers/pdfs/2006/frimanTMI06.pdf Friman, O., Farneback, G., Westin CF. A Bayesian Approach for Stochastic White Matter Tractography. IEEE Transactions on Medical Imaging, Vol 25, No. 8, Aug. 2006]

[[Image:StochasticGUI.png|thumb|right|200px|<font size=1> Figure 4: Python Stochastic Tractography GUI </font>]]

==Module==
Can be found in: MODULES > PYTHON MODULES > PYTHON STOCHASTIC TRACTOGRAPHY
;Functionality of Python Stochastic Tractography module in Slicer 3.0
* IO:
Module reads files (DWI and ROIs) in nhdr format.
* Smoothing:
One can smooth the DWI data (only Gausian smoothing is supported at this time). We recommend it if the data is noisy.
* Brain Mask:
The Brain mask defines the volume in which the tensor will be computed and the tracts evaluated.
*Diffusion Tensor:
This step calculates the tensor and can output anisotropy indices (FA/Mode/Trace)
*Tractography:
Parameters that need to be adjusted:
:1. The amount of tracts that will be seeded from each voxel (we recommend between 500 and 1000 tracts, depending on the workstation power- 1000 tracts per voxel seeded within the large ROI for high resolution DWI can take a long time to compute).
:2. Maximum tract length (in mm), this can eliminate long, unwanted tracts if the regions for which connection is measured are located close to each other
:3. Step Size: distance between each re-estimation of tensors, usually between 0.5 and 1 mm
:4. Stopping criteria. This can be used on the top of WM mask, to terminate tracts (in case they really want to travel through the CSF, for example).
*Connectivity Map:
This step creates output probability maps.
:1. binary: each voxel is counted only once if at least one fiber pass through it
:2. cumulative: tracts are summed by voxel independently
:3. weighted: tracts are summed by voxel depending on their length

Then, probability maps can be saved as ROIs, and either used directly, or thresholded (at certain probability, step claimed by few publications to remove noise) in slicer to mask and compute average FA, Mode, Trace for entire connection. Diffusion indices can be also weighted by the probability of connection for each voxel.

:* Module documentation can be found here:
:**[[Media:IJdata.tar.gz|Training Dataset]]
:**[[Media:Stochastic_June09.ppt|Training Presentation]]
:**[[Media:Helix.zip|Sample Helix Dataset]]

==Work Accomplished==
[[Image:helix_withsmoothing.png|thumb|right|200px|<font size=1>Figure 6: Stochastic tractography from a single ROI on helix phantom</font>]]
; A - Optimization and testing of stochastic tractography algorithm :
:* Original methodological paper, as well as our first attempts to use the algorithm (CC+ and matlab scripts) have been done on old "NAMIC" 1.5T LSDI data ([https://portal.nbirn.net:443/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=gotoDirectory&CPq_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FHarvard%2Farchive_morph&up=CPq&JavaScript=enabled Structural MRI] and [https://portal.nbirn.net:443/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=gotoDirectory&up=7li&7li_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FHarvard%2Farchive_diffusion&JavaScript=enabled DTI data]).
:* Tri worked hard on making sure algorithm works on new high resolution 3T data (available here: [https://portal.nbirn.net/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=moveUpDir&gv7_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FPNL%2F3T_strct_dti_fmri%2Fcase01017&up=gv7&JavaScript=enabled|PNL 3T Data]).
:* Tests have been done also on the spiral diffusion phantom, to make sure diffusion directions and scanner coordinates are handled properly by the algorithm (Figure 6).

; B - Clinical Applications

:* Algorithm was used to trace and analyze anterior limb of the internal capsule on 1.5T data. It generated reacher representation of frontal fiber projections, it also turned out to be more sensitive to group differences in white matter integrity that conventional deterministic tractography (see Figure 6).
:* Algorithm was also used to trace smaller white matter fiber tracts, such as Cingulum, Fornix, Uncinate Fasciculus, Arcuate Fasciculus on 3T "Santa Fe" dataset (http://www.na-mic.org/Wiki/index.php/SanteFe.Tractography.Conference)

; C - References

:* [http://www.na-mic.org/Wiki/index.php/Image:IC_posternew.png Shenton, M.E., Ngo, T., Rosenberger, G., Westin, C.F., Levitt, J.J., McCarley, R.W., Kubicki, M. Study of Thalamo-Cortical White Matter Fiber Tract Projections in Schizophrenia Using Diffusion Stochastic Tractography. Poster presented at the 46th Meeting of the American College of Neuropsychopharmacology, Boca Raton, FL, December 2007.]
:* [http://www.na-mic.org/Wiki/index.php/Media:dougt_poster.pdf Terry DP, Rausch AC, Alvarado JL, Melonakos ED, Markant D, Westin CF, Kikinis R, von Siebenthal J, Shenton ME, Kubicki M. White Matter Properties of Emotion Related Connections in Schizophrenia. Poster presented at the 2009 Mysell Poster Day, Dept. of Psychiatry, Harvard Medical School, April 2009]
:* [http://www.na-mic.org/Wiki/index.php/Media:Jorge_poster.pdf Alvarado JL, Terry DP, Markant D, Ngo T, Kikinis R, Westin CF, McCarley RW, Shenton ME, Kubicki M. Study of Language-Related White Matter Tract Connections in Schizophrenia using Diffusion Stochastic Tractography. Poster presented at the 2009 Mysell Poster Day, Dept. of Psychiatry, Harvard Medical School, April 2009]
:* Melonakos ED, Shenton ME, Markant D, Alvarado J, Westin CF, Kubicki M. White Matter Properties of Orbitofrontal Connections in Schizophrenia. Poster being presented at the 64th Meeting of the Society of Biological Psychiatry. Vancouver, BC. May 2009.
:* Kubicki, M. Khan, U., Bobrow, L., O'Donnell, L. Pieper, S. Westin, CF., Shenton, ME. New Methods for Assessing Whole Brain DTI Abnormalities in Schizophrenia. Presentation given at the International Congress of World Psychiatric Association. Florence, Italy. April 2009.
:* Kubicki, M., Markant, D., Ngo, T., Westin, CF., McCarley, RW., Shenton, ME. Study of Language Related White Matter Fiber Tract Projections in Schizophrenia Using Diffusion Stochastic Tractography. Presentation given at the International Congress of World Psychiatric Association. Florence, Italy. April 2009.

== Work in Progress ==

; A - Optimization and Testing of stochastic tractography module :
:* Julien is testing now separate components of work flow, including the masks, and impact of their precision on stochastic output, as well as impact of number of seeding points on tractography results.
:* At the same time, we are testing the module on a Max Plank dataset of Anna Rotarska-Jagiela by looking at the connections through the corpus callosum of the left auditory cortex to the right auditory cortex as defined by fMRI activation maps as well as running this data for the tractography comparison project.
:* We are discussing possibility of adding project specific functionality to module, such as third ROI to guide tractography, nonlinear registration button for merging fMRI and DTI data.

; B - Related Clinical Projects

;* Arcuate Fasciculus Extraction Project
[[Image:STArcuate.jpg|thumb|right|200px|<font size=1>Figure 7: The arcuate fasciculus including seed, midpoint and target ROI's.</font>]]
We have started the project of investigating Arcuate Fasciculus using Stochastic Tractography (Figure 7). This structure is especially important in both VCFS and schizophrenia, as it connects language related areas (Brocka and Wernicke's), and is involved in language processing quite disturbed in schizophrenia patients. It also can not be reliably traced using deterministic tractography.
:Project involves:
:* Whole brain segmentation, and automatic extraction of regions interconnected by Arcuate Fasciculus (Inferior frontal and Superior Temporal Gyri).
:* White matter segmentation, in order to prevent algorithm from traveling through the ventricles, where diffusivity is high.
:* Non-linear registration of labelmaps to the DTI space.
:* Seeding tracts. We have piloted it using 5000 seeds per voxel, however it is quite time consuming running it on even most powerful computers in the lab, so we have experimented with smaller number of seeds per voxel. We tested 1000 seeds, which gave virtually identical results.
:* Extracting path of interest, and calculating FA along the path for group comparison. Presentation of previous results for 7 schizophrenics and 12 control subjects, can be found here: [[Media:NAMIC_AHM_Arcuate.ppt|Progress Report Presentation]].
:* Results presentation and paper submission. Abstract was submitted and accepted for presentation at World Biological Psychiatry Symposium in Venice, Italy. Paper is in preparation.

;* Semantic Network Connectivity Project
We use combination of fMRI and DTI data to define and characterize functional and anatomical connectivity within the semantic processing network in schizophrenia.
:Project involves:
:* fMRI data analysis and identification of functional nodes involved in semantic processing in healthy controls and seubjects with schizophrenia
:* Analysis of functional connectivity (using FSL) between nodes of semantic network
:* Whole brain Voxel Based analysis of DTI data in same population
:* Use of stochastic tractography to identify connections between functional nodes
:* Correlational analysis involving anatomical and functional connectivity data.
:* Results presentation and paper submission. Data was presented at HBM conference in 2007, paper has been submitted to HBM.
[[Image:Anna.png|thumb|right|200px|<font size=1>Figure 8: Max Plank data showing tracts through the corpus connecting 2 cortical ROIs defined by fMRI activations.</font>]]

;* Study of Default Network
We have started collaboration with Department of Neurophysiology Max Planck Institute in Frankfurt (Anna Rotarska contact person). They have dataset containing DTI and resting state fMRI in schizophrenia, and want to use stochastic tractography to measure integrity of anatomical connections within the default network in schizophrenia. Their fMRI data has been co-registered with anatomical scans, and has been put it into DTI space. Figure 8 shows pilot data of the white matter connections through the corpus callosum between the left and the right auditory cortex ROIs as defined by fMRI data. This data will be also used to test robustness of our module, since data was collected on a different scanner (3T Philips).

;* Tractography Comparison Project
We are also working on a [http://www.na-mic.org/Wiki/index.php/SanteFe.Tractography.Conference tractography comparison project]dataset, where we apply stochastic tractography to phantom, as well as test dataset.

===Staffing Plan===

* Sylvain and Yogesh are the DBP resources charged with adapting the tools in the NA-MIC Kit to the DBP needs
* Doug Markant, our NAMIC RA has left the lab, and now Doug Terry, is a new NAMIC RA.
* Julien is our new NAMIC software engineer. He is responsible for improving the STM (Stochastic Tractography Module), and making sure software works with STM compliant datasets.
[[Link Progress| Development Progress]]
* Polina is the algorithm core contact
* Brad is the engineering core contact

===Schedule===

* '''10/2007''' - Optimization of Stochastic Tractography algorythm for 1.5T data.
* '''10/2007''' - Algorythm testing on Santa Fe data set and diffusion phantom.
* '''06/2008''' - Optimization of Stochastic Tractography algorythm for 3T data.
* '''11/2008''' - Slicer 3 module prototype using python.
* '''12/2008''' - Slicer 3 module official release
* '''12/2008''' - Documentation and packaging for dissemination.
* '''12/2008''' - Arcuate Fasciculus results.
* '''01/2009''' - Arcuate Fasciuclus first draft of the paper.
* '''05/2009''' - Distortion correction and nonlinear registration added to the module
* '''05/2009''' - Symposium on tractography, including stochastic methods at World Biological Psychiatry Symposium in Florence, Italy.
* '''05/2009''' - Presentation of Arcuate Fasciculus findings at World Biological Psychiatry Symposium in Florence, Italy.

===Team and Institute===
*PI: Marek Kubicki (kubicki at bwh.harvard.edu)
*DBP2 Investigators: Sylvain Bouix, Yogesh Rathi, Julien de Siebenthal
*NA-MIC Engineering Contact: Brad Davis, Kitware
*NA-MIC Algorithms Contact: Polina Gollard, MIT

===Publications===

''In print''

*[http://www.na-mic.org/pages/Special:Publications?text=Kubicki+AND+Westin+AND+DTI&submit=Search&words=all&title=checked&keywords=checked&authors=checked&abstract=checked&sponsors=checked&searchbytag=checked| NA-MIC Publications Database - Clinical Applications]

*[http://www.na-mic.org/pages/Special:Publications?text=Ngo+AND+Golland&submit=Search&words=all&title=checked&keywords=checked&authors=checked&abstract=checked&sponsors=checked&searchbytag=checked| NA-MIC Publications Database - Algorithms Development]

[[Category: Schizophrenia]] [[Category: Diffusion MRI]] [[Category: Segmentation]]

File:StochasticGUI.PNG

2009-06-03T18:09:44Z

Dougt:

File:StochasticGUI.png

2009-06-03T18:05:46Z

Dougt:

DBP2:Harvard:Brain Segmentation Roadmap

2009-06-03T17:48:02Z

Dougt: /* Module */

Back to [[NA-MIC_Internal_Collaborations|NA-MIC Collaborations]], [[DBP2:Harvard|Harvard DBP 2]]
__NOTOC__
=Stochastic Tractography for VCFS=
== Roadmap ==

The main goal of this project is to develop end-to-end application that would be used to characterize anatomical connectivity abnormalities in the brain of patients with velocardiofacial syndrome (VCFS), and to link this information with deficits in schizophrenia. This page describes the technology roadmap for stochastic tractography, using newly acquired 3T data, NAMIC tools and slicer 3.

== Algorithm ==

[[Image:IC_sto_new.png|thumb|right|200px|<font size=1> Figure 1: Comparison of deterministic and stochastic tractography algorithms</font>]]
; A-Description
* Most tractography methods estimate fibers by tracing the maximum direction of diffusion. A limitation of this approach is that, in practice, several factors introduce uncertainty in the tracking procedure, including, noise, splitting and crossing fibers, head motion and image artifacts. To address this uncertainty, stochastic tractography methods have been developed to quantify the uncertainty associated with estimated fibers (Bjornemo et al., 2002). Method uses a propagation model based on stochastics and regularization, which allows paths originating at one point to branch and return a probability distribution of possible paths. The method utilizes principles of a statistical Monte Carlo method called Sequential Importance Sampling and Resampling (SISR). Based on probability functions, using a sequential importance sampling technique ([http://lmi.bwh.harvard.edu/papers/pdfs/2002/bjornemoMICCAI02.pdf Bjornemo et al., 2002]), one can generate thousands of fibers starting in the same point by sequentially drawing random step directions. This gives a very rich model of the fiber distribution, as contrasted with single fibers produced by conventional tractography methods. Moreover, from a large number of sampled paths, probability maps can be generated, providing better estimates of connectivity between several anatomical locations. A comparison of the algorithms can be seen here. (Figure 1)

[[Image:StochasticPic.PNG|thumb|right|200px|<font size=1>Figure 2: Stochastic tractography of uncinate fasciculis on anatomical data (left) and cingulum bungle on fMRI scan (right)</font>]]
; B-Possible Applications
* Since diffusion direction uncertainty within the gray matter is quite significant; principal diffusion direction approaches usually do not work for tracking between two gray matter regions. Thus if one requires finding connections between a priori selected anatomical gray matter regions, defined either by anatomical segmentations (in case of using structural ROI data), or functional activations (in case of megring DTI with fMRI), stochastic tractography seems to be the method of choice. Here is an example of this application to anatomical data (Figure 2, left image) and to fMRI data (Figure 2, right image).

* Stochastic Tractography is also comparable, if not better, in defining large white matter fiber bundles, especially those traveling through white matter regions characterized by increased diffusion uncertainty (fiber crossings). Example of such application to internal capsule. (Figure 3)

[[Image:IC-comp-new.png|thumb|right|200px|<font size=1>Figure 3: Streamline vs. stochastic tractography of the Internal Capsule</font>]]
; C-References

* [http://lmi.bwh.harvard.edu/papers/pdfs/2002/bjornemoMICCAI02.pdf Björnemo M, Brun A, Kikinis R, Westin CF. Regularized stochastic white matter tractography using diffusion tensor MRI. In Fifth International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI'02). Tokyo, Japan, 2002;435-442.]
* [http://lmi.bwh.harvard.edu/papers/pdfs/2006/frimanTMI06.pdf Friman, O., Farneback, G., Westin CF. A Bayesian Approach for Stochastic White Matter Tractography. IEEE Transactions on Medical Imaging, Vol 25, No. 8, Aug. 2006]

[[Image:stochastic_panel.png|thumb|right|200px|<font size=1> Figure 4: Python Stochastic Tractography GUI </font>]]

==Module==
Can be found in: MODULES > PYTHON MODULES > PYTHON STOCHASTIC TRACTOGRAPHY
;Functionality of Python Stochastic Tractography module in Slicer 3.0
* IO:
Module reads files (DWI and ROIs) in nhdr format.
* Smoothing:
One can smooth the DWI data (only Gausian smoothing is supported at this time). We recommend it if the data is noisy.
* Brain Mask:
The Brain mask defines the volume in which the tensor will be computed and the tracts evaluated.
*Diffusion Tensor:
This step calculates the tensor and can output anisotropy indices (FA/Mode/Trace)
*Tractography:
Parameters that need to be adjusted:
:1. The amount of tracts that will be seeded from each voxel (we recommend between 500 and 1000 tracts, depending on the workstation power- 1000 tracts per voxel seeded within the large ROI for high resolution DWI can take a long time to compute).
:2. Maximum tract length (in mm), this can eliminate long, unwanted tracts if the regions for which connection is measured are located close to each other
:3. Step Size: distance between each re-estimation of tensors, usually between 0.5 and 1 mm
:4. Stopping criteria. This can be used on the top of WM mask, to terminate tracts (in case they really want to travel through the CSF, for example).
*Connectivity Map:
This step creates output probability maps.
:1. binary: each voxel is counted only once if at least one fiber pass through it
:2. cumulative: tracts are summed by voxel independently
:3. weighted: tracts are summed by voxel depending on their length

Then, probability maps can be saved as ROIs, and either used directly, or thresholded (at certain probability, step claimed by few publications to remove noise) in slicer to mask and compute average FA, Mode, Trace for entire connection. Diffusion indices can be also weighted by the probability of connection for each voxel.

:* Module documentation can be found here:
:**[[Media:IJdata.tar.gz|Training Dataset]]
:**[[Media:Stochastic_June09.ppt|Training Presentation]]
:**[[Media:Helix.zip|Sample Helix Dataset]]

==Work Accomplished==
[[Image:helix_withsmoothing.png|thumb|right|200px|<font size=1>Figure 6: Stochastic tractography from a single ROI on helix phantom</font>]]
; A - Optimization and testing of stochastic tractography algorithm :
:* Original methodological paper, as well as our first attempts to use the algorithm (CC+ and matlab scripts) have been done on old "NAMIC" 1.5T LSDI data ([https://portal.nbirn.net:443/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=gotoDirectory&CPq_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FHarvard%2Farchive_morph&up=CPq&JavaScript=enabled Structural MRI] and [https://portal.nbirn.net:443/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=gotoDirectory&up=7li&7li_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FHarvard%2Farchive_diffusion&JavaScript=enabled DTI data]).
:* Tri worked hard on making sure algorithm works on new high resolution 3T data (available here: [https://portal.nbirn.net/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=moveUpDir&gv7_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FPNL%2F3T_strct_dti_fmri%2Fcase01017&up=gv7&JavaScript=enabled|PNL 3T Data]).
:* Tests have been done also on the spiral diffusion phantom, to make sure diffusion directions and scanner coordinates are handled properly by the algorithm (Figure 6).

; B - Clinical Applications

:* Algorithm was used to trace and analyze anterior limb of the internal capsule on 1.5T data. It generated reacher representation of frontal fiber projections, it also turned out to be more sensitive to group differences in white matter integrity that conventional deterministic tractography (see Figure 6).
:* Algorithm was also used to trace smaller white matter fiber tracts, such as Cingulum, Fornix, Uncinate Fasciculus, Arcuate Fasciculus on 3T "Santa Fe" dataset (http://www.na-mic.org/Wiki/index.php/SanteFe.Tractography.Conference)

; C - References

:* [http://www.na-mic.org/Wiki/index.php/Image:IC_posternew.png Shenton, M.E., Ngo, T., Rosenberger, G., Westin, C.F., Levitt, J.J., McCarley, R.W., Kubicki, M. Study of Thalamo-Cortical White Matter Fiber Tract Projections in Schizophrenia Using Diffusion Stochastic Tractography. Poster presented at the 46th Meeting of the American College of Neuropsychopharmacology, Boca Raton, FL, December 2007.]
:* [http://www.na-mic.org/Wiki/index.php/Media:dougt_poster.pdf Terry DP, Rausch AC, Alvarado JL, Melonakos ED, Markant D, Westin CF, Kikinis R, von Siebenthal J, Shenton ME, Kubicki M. White Matter Properties of Emotion Related Connections in Schizophrenia. Poster presented at the 2009 Mysell Poster Day, Dept. of Psychiatry, Harvard Medical School, April 2009]
:* [http://www.na-mic.org/Wiki/index.php/Media:Jorge_poster.pdf Alvarado JL, Terry DP, Markant D, Ngo T, Kikinis R, Westin CF, McCarley RW, Shenton ME, Kubicki M. Study of Language-Related White Matter Tract Connections in Schizophrenia using Diffusion Stochastic Tractography. Poster presented at the 2009 Mysell Poster Day, Dept. of Psychiatry, Harvard Medical School, April 2009]
:* Melonakos ED, Shenton ME, Markant D, Alvarado J, Westin CF, Kubicki M. White Matter Properties of Orbitofrontal Connections in Schizophrenia. Poster being presented at the 64th Meeting of the Society of Biological Psychiatry. Vancouver, BC. May 2009.
:* Kubicki, M. Khan, U., Bobrow, L., O'Donnell, L. Pieper, S. Westin, CF., Shenton, ME. New Methods for Assessing Whole Brain DTI Abnormalities in Schizophrenia. Presentation given at the International Congress of World Psychiatric Association. Florence, Italy. April 2009.
:* Kubicki, M., Markant, D., Ngo, T., Westin, CF., McCarley, RW., Shenton, ME. Study of Language Related White Matter Fiber Tract Projections in Schizophrenia Using Diffusion Stochastic Tractography. Presentation given at the International Congress of World Psychiatric Association. Florence, Italy. April 2009.

== Work in Progress ==

; A - Optimization and Testing of stochastic tractography module :
:* Julien is testing now separate components of work flow, including the masks, and impact of their precision on stochastic output, as well as impact of number of seeding points on tractography results.
:* At the same time, we are testing the module on a Max Plank dataset of Anna Rotarska-Jagiela by looking at the connections through the corpus callosum of the left auditory cortex to the right auditory cortex as defined by fMRI activation maps as well as running this data for the tractography comparison project.
:* We are discussing possibility of adding project specific functionality to module, such as third ROI to guide tractography, nonlinear registration button for merging fMRI and DTI data.

; B - Related Clinical Projects

;* Arcuate Fasciculus Extraction Project
[[Image:STArcuate.jpg|thumb|right|200px|<font size=1>Figure 7: The arcuate fasciculus including seed, midpoint and target ROI's.</font>]]
We have started the project of investigating Arcuate Fasciculus using Stochastic Tractography (Figure 7). This structure is especially important in both VCFS and schizophrenia, as it connects language related areas (Brocka and Wernicke's), and is involved in language processing quite disturbed in schizophrenia patients. It also can not be reliably traced using deterministic tractography.
:Project involves:
:* Whole brain segmentation, and automatic extraction of regions interconnected by Arcuate Fasciculus (Inferior frontal and Superior Temporal Gyri).
:* White matter segmentation, in order to prevent algorithm from traveling through the ventricles, where diffusivity is high.
:* Non-linear registration of labelmaps to the DTI space.
:* Seeding tracts. We have piloted it using 5000 seeds per voxel, however it is quite time consuming running it on even most powerful computers in the lab, so we have experimented with smaller number of seeds per voxel. We tested 1000 seeds, which gave virtually identical results.
:* Extracting path of interest, and calculating FA along the path for group comparison. Presentation of previous results for 7 schizophrenics and 12 control subjects, can be found here: [[Media:NAMIC_AHM_Arcuate.ppt|Progress Report Presentation]].
:* Results presentation and paper submission. Abstract was submitted and accepted for presentation at World Biological Psychiatry Symposium in Venice, Italy. Paper is in preparation.

;* Semantic Network Connectivity Project
We use combination of fMRI and DTI data to define and characterize functional and anatomical connectivity within the semantic processing network in schizophrenia.
:Project involves:
:* fMRI data analysis and identification of functional nodes involved in semantic processing in healthy controls and seubjects with schizophrenia
:* Analysis of functional connectivity (using FSL) between nodes of semantic network
:* Whole brain Voxel Based analysis of DTI data in same population
:* Use of stochastic tractography to identify connections between functional nodes
:* Correlational analysis involving anatomical and functional connectivity data.
:* Results presentation and paper submission. Data was presented at HBM conference in 2007, paper has been submitted to HBM.
[[Image:Anna.png|thumb|right|200px|<font size=1>Figure 8: Max Plank data showing tracts through the corpus connecting 2 cortical ROIs defined by fMRI activations.</font>]]

;* Study of Default Network
We have started collaboration with Department of Neurophysiology Max Planck Institute in Frankfurt (Anna Rotarska contact person). They have dataset containing DTI and resting state fMRI in schizophrenia, and want to use stochastic tractography to measure integrity of anatomical connections within the default network in schizophrenia. Their fMRI data has been co-registered with anatomical scans, and has been put it into DTI space. Figure 8 shows pilot data of the white matter connections through the corpus callosum between the left and the right auditory cortex ROIs as defined by fMRI data. This data will be also used to test robustness of our module, since data was collected on a different scanner (3T Philips).

;* Tractography Comparison Project
We are also working on a [http://www.na-mic.org/Wiki/index.php/SanteFe.Tractography.Conference tractography comparison project]dataset, where we apply stochastic tractography to phantom, as well as test dataset.

===Staffing Plan===

* Sylvain and Yogesh are the DBP resources charged with adapting the tools in the NA-MIC Kit to the DBP needs
* Doug Markant, our NAMIC RA has left the lab, and now Doug Terry, is a new NAMIC RA.
* Julien is our new NAMIC software engineer. He is responsible for improving the STM (Stochastic Tractography Module), and making sure software works with STM compliant datasets.
[[Link Progress| Development Progress]]
* Polina is the algorithm core contact
* Brad is the engineering core contact

===Schedule===

* '''10/2007''' - Optimization of Stochastic Tractography algorythm for 1.5T data.
* '''10/2007''' - Algorythm testing on Santa Fe data set and diffusion phantom.
* '''06/2008''' - Optimization of Stochastic Tractography algorythm for 3T data.
* '''11/2008''' - Slicer 3 module prototype using python.
* '''12/2008''' - Slicer 3 module official release
* '''12/2008''' - Documentation and packaging for dissemination.
* '''12/2008''' - Arcuate Fasciculus results.
* '''01/2009''' - Arcuate Fasciuclus first draft of the paper.
* '''05/2009''' - Distortion correction and nonlinear registration added to the module
* '''05/2009''' - Symposium on tractography, including stochastic methods at World Biological Psychiatry Symposium in Florence, Italy.
* '''05/2009''' - Presentation of Arcuate Fasciculus findings at World Biological Psychiatry Symposium in Florence, Italy.

===Team and Institute===
*PI: Marek Kubicki (kubicki at bwh.harvard.edu)
*DBP2 Investigators: Sylvain Bouix, Yogesh Rathi, Julien de Siebenthal
*NA-MIC Engineering Contact: Brad Davis, Kitware
*NA-MIC Algorithms Contact: Polina Gollard, MIT

===Publications===

''In print''

*[http://www.na-mic.org/pages/Special:Publications?text=Kubicki+AND+Westin+AND+DTI&submit=Search&words=all&title=checked&keywords=checked&authors=checked&abstract=checked&sponsors=checked&searchbytag=checked| NA-MIC Publications Database - Clinical Applications]

*[http://www.na-mic.org/pages/Special:Publications?text=Ngo+AND+Golland&submit=Search&words=all&title=checked&keywords=checked&authors=checked&abstract=checked&sponsors=checked&searchbytag=checked| NA-MIC Publications Database - Algorithms Development]

[[Category: Schizophrenia]] [[Category: Diffusion MRI]] [[Category: Segmentation]]

Python Stochastic Tractography Tutorial

2009-06-03T17:38:16Z

Dougt:

{|
|[[Image:NAMIC-SLC.jpg|thumb|320px|Return to [[2009_Winter_Project_Week|Project Week Main Page]] ]]
|[[Image:Cc_01.png|thumb|340px|Model of the Corpus Callosum from Stochastic Tractography (coronal) ]]
|[[Image:Cc_02.png|thumb|340px|Model of the Corpus Callosum from Stochastic Tractography (sagittal) ]]
|[[Image:helix_withsmoothing.png|thumb|220px|Stochastic Tractography on Phantom]]
|}

__NOTOC__

===Key Investigators===
*PI: Marek Kubicki (kubicki at bwh.harvard.edu)
*DBP2 Investigators: Sylvain Bouix, Yogesh Rathi, Julien de Siebenthal, RA Doug Terry
*NA-MIC Engineering Contact: Brad Davis, Kitware
*NA-MIC Algorithms Contact: Polina Gollard, MIT

Go to [http://www.slicer.org/slicerWiki/index.php/Modules:StochasticTractography-Documentation-3.4 Slicer 3.4 module documentation] for more information

===Tutorial===
*[[Media:IJdata.tar.gz|Training Dataset]]
*[[Media:Stochastic_June09.ppt|Training Presentation]]
*[[Media:Helix.zip|Sample Helix Dataset]]

__NOTOC__
<div style="margin: 20px;">
<div style="width: 40%; float: left; padding-right: 3%;">
<h1>Objective</h1>
To address the uncertainty of deterministic tractography methods, stochastic tractography methods have been developed to quantify the uncertainty associated with estimated fibers (Bjornemo et al., 2002). Method uses a propagation model based on stochastics and regularization, which allows paths originating at one point to branch and return a probability distribution of possible paths. One can generate thousands of fibers starting in the same point by sequentially drawing random step directions. This gives a very rich model of the fiber distribution, as contrasted with single fibers produced by conventional tractography methods. Moreover, from a large number of sampled paths, probability maps can be generated, providing better estimates of connectivity between several anatomical locations.
</div>

<div style="width: 27%; float: left; padding-right: 3%;">
<h1>Approach, Plan</h1>
This tutorial will walk users through the multiple stages of the python stochastic tractography module, where the inputs are a DWI and a region of interest labelmap. This will include:
# Loading the Volumes
# Extracting DWI and getting Baseline DWI threshold levels
# Using features of the module including smoothing, brain mask generation, and other settings.
# Generating a Connectivity Map, which is able to be thresholded and customized.
</div>

<div style="width: 27%; float: left;">
<h1>Progress</h1>
*A tutorial has been developed and data is available for tutorial.
*We have started the project of investigating Arcuate Fasciculus using Stochastic Tractography. This structure is especially important in both VCFS and schizophrenia, as it connects language related areas (Brocka and Wernicke's), and is involved in language processing quite disturbed in schizophrenia patients. It also can not be reliably traced using deterministic tractography.
</div>

<br style="clear: both;" />
</div>

File:Stochastic June09.ppt

2009-06-03T17:37:25Z

Dougt:

DBP2:Harvard:Brain Segmentation Roadmap

2009-04-16T18:27:33Z

Dougt: /* Work in Progress */

Back to [[NA-MIC_Internal_Collaborations|NA-MIC Collaborations]], [[DBP2:Harvard|Harvard DBP 2]]
__NOTOC__
=Stochastic Tractography for VCFS=
== Roadmap ==

The main goal of this project is to develop end-to-end application that would be used to characterize anatomical connectivity abnormalities in the brain of patients with velocardiofacial syndrome (VCFS), and to link this information with deficits in schizophrenia. This page describes the technology roadmap for stochastic tractography, using newly acquired 3T data, NAMIC tools and slicer 3.

== Algorithm ==

[[Image:IC_sto_new.png|thumb|right|200px|<font size=1> Figure 1: Comparison of deterministic and stochastic tractography algorithms</font>]]
; A-Description
* Most tractography methods estimate fibers by tracing the maximum direction of diffusion. A limitation of this approach is that, in practice, several factors introduce uncertainty in the tracking procedure, including, noise, splitting and crossing fibers, head motion and image artifacts. To address this uncertainty, stochastic tractography methods have been developed to quantify the uncertainty associated with estimated fibers (Bjornemo et al., 2002). Method uses a propagation model based on stochastics and regularization, which allows paths originating at one point to branch and return a probability distribution of possible paths. The method utilizes principles of a statistical Monte Carlo method called Sequential Importance Sampling and Resampling (SISR). Based on probability functions, using a sequential importance sampling technique ([http://lmi.bwh.harvard.edu/papers/pdfs/2002/bjornemoMICCAI02.pdf Bjornemo et al., 2002]), one can generate thousands of fibers starting in the same point by sequentially drawing random step directions. This gives a very rich model of the fiber distribution, as contrasted with single fibers produced by conventional tractography methods. Moreover, from a large number of sampled paths, probability maps can be generated, providing better estimates of connectivity between several anatomical locations. A comparison of the algorithms can be seen here. (Figure 1)

[[Image:StochasticPic.PNG|thumb|right|200px|<font size=1>Figure 2: Stochastic Tractography of uncinate fasciculis on anatomical data (left) and cingulum bungle on fMRI scan (right)</font>]]
; B-Possible Applications
* Since diffusion direction uncertainty within the gray matter is quite significant; principal diffusion direction approaches usually do not work for tracking between two gray matter regions. Thus if one requires finding connections between a priori selected anatomical gray matter regions, defined either by anatomical segmentations (in case of using structural ROI data), or functional activations (in case of megring DTI with fMRI), stochastic tractography seems to be the method of choice. Here is an example of this application to anatomical data (Figure 2, left image) and to fMRI data (Figure 2, right image).

* Stochastic Tractography is also comparable, if not better, in defining large white matter fiber bundles, especially those traveling through white matter regions characterized by increased diffusion uncertainty (fiber crossings). Example of such application to internal capsule. (Figure 3)

[[Image:IC-comp-new.png|thumb|right|200px|<font size=1>Figure 3: Streamline vs. Stochastic Tractography of the Internal Capsule</font>]]
; C-References

* [http://lmi.bwh.harvard.edu/papers/pdfs/2002/bjornemoMICCAI02.pdf Björnemo M, Brun A, Kikinis R, Westin CF. Regularized stochastic white matter tractography using diffusion tensor MRI. In Fifth International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI'02). Tokyo, Japan, 2002;435-442.]
* [http://lmi.bwh.harvard.edu/papers/pdfs/2006/frimanTMI06.pdf Friman, O., Farneback, G., Westin CF. A Bayesian Approach for Stochastic White Matter Tractography. IEEE Transactions on Medical Imaging, Vol 25, No. 8, Aug. 2006]

[[Image:stoch_menu.png|thumb|right|200px|<font size=1> Figure 4: Python Stochastic Tractography GUI </font>]]

==Module==
Can be found in: MODULES > PYTHON MODULES > PYTHON STOCHASTIC TRACTOGRAPHY
;Functionality of Python Stochastic Tractography module in Slicer 3.0
* IO:
Module reads files (DWI and ROIs) in nhdr format.
* Smoothing:
One can smooth the DWI data (only Gausian smoothing is supported at this time). We recommend it if the data is noisy.
* Brain Mask:
The Brain mask defines the volume in which the tensor will be computed and the tracts evaluated.
*Diffusion Tensor:
This step calculates the tensor and can output anisotropy indices (FA/Mode/Trace)
*Tractography:
Parameters that need to be adjusted:
:1. The amount of tracts that will be seeded from each voxel (we recommend between 500 and 1000 tracts, depending on the workstation power- 1000 tracts per voxel seeded within the large ROI for high resolution DWI can take a long time to compute).
:2. Maximum tract length (in mm), this can eliminate long, unwanted tracts if the regions for which connection is measured are located close to each other [[Image:stoch_menu2a.png|thumb|right|200px|<font size=1> Figure 5: Python Stochastic Tractography GUI, part 2 </font>]]
:3. Step Size: distance between each re-estimation of tensors, usually between 0.5 and 1 mm
:4. Stopping criteria. This can be used on the top of WM mask, to terminate tracts (in case they really want to travel through the CSF, for example).
*Connectivity Map:
This step creates output probability maps.
:1. binary: each voxel is counted only once if at least one fiber pass through it
:2. cumulative: tracts are summed by voxel independently
:3. weighted: tracts are summed by voxel depending on their length

Then, probability maps can be saved as ROIs, and either used directly, or thresholded (at certain probability, step claimed by few publications to remove noise) in slicer to mask and compute average FA, Mode, Trace for entire connection. Diffusion indices can be also weighted by the probability of connection for each voxel.

:* Module documentation can be found here:
:**[[Media:IJdata.tar.gz|Training Dataset]]
:**[[Media:PNLStochasictNew.ppt|Training Presentation]]
:**[[Media:Helix.zip|Sample Helix Dataset]]

==Work Accomplished==
[[Image:helix.png|thumb|right|200px|<font size=1>Figure 6: Stochastic Tractography on Phantom</font>]]
; A - Optimization and testing of stochastic tractography algorithm :
:* Original methodological paper, as well as our first attempts to use the algorithm (CC+ and matlab scripts) have been done on old "NAMIC" 1.5T LSDI data ([https://portal.nbirn.net:443/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=gotoDirectory&CPq_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FHarvard%2Farchive_morph&up=CPq&JavaScript=enabled Structural MRI] and [https://portal.nbirn.net:443/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=gotoDirectory&up=7li&7li_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FHarvard%2Farchive_diffusion&JavaScript=enabled DTI data]).
:* Tri worked hard on making sure algorithm works on new high resolution 3T data (available here: [https://portal.nbirn.net/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=moveUpDir&gv7_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FPNL%2F3T_strct_dti_fmri%2Fcase01017&up=gv7&JavaScript=enabled|PNL 3T Data]).
:* Tests have been done also on the spiral diffusion phantom, to make sure diffusion directions and scanner coordinates are handled properly by the algorithm (Figure 6).

; B - Clinical Applications

:* Algorithm was used to trace and analyze anterior limb of the internal capsule on 1.5T data. It generated reacher representation of frontal fiber projections, it also turned out to be more sensitive to group differences in white matter integrity that conventional deterministic tractography (see Figure 6).
:* Algorithm was also used to trace smaller white matter fiber tracts, such as Cingulum, Fornix, Uncinate Fasciculus, Arcuate Fasciculus on 3T "Santa Fe" dataset (http://www.na-mic.org/Wiki/index.php/SanteFe.Tractography.Conference)

; C - References

:* [http://www.na-mic.org/Wiki/index.php/Image:IC_posternew.png Shenton, M.E., Ngo, T., Rosenberger, G., Westin, C.F., Levitt, J.J., McCarley, R.W., Kubicki, M. Study of Thalamo-Cortical White Matter Fiber Tract Projections in Schizophrenia Using Diffusion Stochastic Tractography. Poster presented at the 46th Meeting of the American College of Neuropsychopharmacology, Boca Raton, FL, December 2007.]
:* [http://www.na-mic.org/Wiki/index.php/Media:dougt_poster.pdf Terry DP, Rausch AC, Alvarado JL, Melonakos ED, Markant D, Westin CF, Kikinis R, von Siebenthal J, Shenton ME, Kubicki M. White Matter Properties of Emotion Related Connections in Schizophrenia. Poster presented at the 2009 Mysell Poster Day, Dept. of Psychiatry, Harvard Medical School, April 2009]
:* [http://www.na-mic.org/Wiki/index.php/Media:Jorge_poster.pdf Alvarado JL, Terry DP, Markant D, Ngo T, Kikinis R, Westin CF, McCarley RW, Shenton ME, Kubicki M. Study of Language-Related White Matter Tract Connections in Schizophrenia using Diffusion Stochastic Tractography. Poster presented at the 2009 Mysell Poster Day, Dept. of Psychiatry, Harvard Medical School, April 2009]
:* Melonakos ED, Shenton ME, Markant D, Alvarado J, Westin CF, Kubicki M. White Matter Properties of Orbitofrontal Connections in Schizophrenia. Poster being presented at the 64th Meeting of the Society of Biological Psychiatry. Vancouver, BC. May 2009.
:* Kubicki, M. Khan, U., Bobrow, L., O'Donnell, L. Pieper, S. Westin, CF., Shenton, ME. New Methods for Assessing Whole Brain DTI Abnormalities in Schizophrenia. Presentation given at the International Congress of World Psychiatric Association. Florence, Italy. April 2009.
:* Kubicki, M., Markant, D., Ngo, T., Westin, CF., McCarley, RW., Shenton, ME. Study of Language Related White Matter Fiber Tract Projections in Schizophrenia Using Diffusion Stochastic Tractography. Presentation given at the International Congress of World Psychiatric Association. Florence, Italy. April 2009.

== Work in Progress ==

; A - Optimization and Testing of stochastic tractography module :
:* Julien is testing now separate components of work flow, including the masks, and impact of their precision on stochastic output, as well as impact of number of seeding points on tractography results.
:* At the same time, we are testing the module on a Max Plank dataset of Anna Rotarska-Jagiela by looking at the connections through the corpus callosum of the left auditory cortex to the right auditory cortex as defined by fMRI activation maps as well as running this data for the tractography comparison project.
:* We are discussing possibility of adding project specific functionality to module, such as third ROI to guide tractography, nonlinear registration button for merging fMRI and DTI data.

; B - Related Clinical Projects

;* Arcuate Fasciculus Extraction Project
[[Image:STArcuate.jpg|thumb|right|200px|<font size=1>Figure 7: The arcuate fasciculus including seed, midpoint and target ROI's.</font>]]
We have started the project of investigating Arcuate Fasciculus using Stochastic Tractography (Figure 7). This structure is especially important in both VCFS and schizophrenia, as it connects language related areas (Brocka and Wernicke's), and is involved in language processing quite disturbed in schizophrenia patients. It also can not be reliably traced using deterministic tractography.
:Project involves:
:* Whole brain segmentation, and automatic extraction of regions interconnected by Arcuate Fasciculus (Inferior frontal and Superior Temporal Gyri).
:* White matter segmentation, in order to prevent algorithm from traveling through the ventricles, where diffusivity is high.
:* Non-linear registration of labelmaps to the DTI space.
:* Seeding tracts. We have piloted it using 5000 seeds per voxel, however it is quite time consuming running it on even most powerful computers in the lab, so we have experimented with smaller number of seeds per voxel. We tested 1000 seeds, which gave virtually identical results.
:* Extracting path of interest, and calculating FA along the path for group comparison. Presentation of previous results for 7 schizophrenics and 12 control subjects, can be found here: [[Media:NAMIC_AHM_Arcuate.ppt|Progress Report Presentation]].
:* Results presentation and paper submission. Abstract was submitted and accepted for presentation at World Biological Psychiatry Symposium in Venice, Italy. Paper is in preparation.

;* Semantic Network Connectivity Project
We use combination of fMRI and DTI data to define and characterize functional and anatomical connectivity within the semantic processing network in schizophrenia.
:Project involves:
:* fMRI data analysis and identification of functional nodes involved in semantic processing in healthy controls and seubjects with schizophrenia
:* Analysis of functional connectivity (using FSL) between nodes of semantic network
:* Whole brain Voxel Based analysis of DTI data in same population
:* Use of stochastic tractography to identify connections between functional nodes
:* Correlational analysis involving anatomical and functional connectivity data.
:* Results presentation and paper submission. Data was presented at HBM conference in 2007, paper has been submitted to HBM.
[[Image:Anna.png|thumb|right|200px|<font size=1>Figure 8: Max Plank data showing tracts through the corpus connecting 2 cortical ROIs defined by fMRI activations.</font>]]

;* Study of Default Network
We have started collaboration with Department of Neurophysiology Max Planck Institute in Frankfurt (Anna Rotarska contact person). They have dataset containing DTI and resting state fMRI in schizophrenia, and want to use stochastic tractography to measure integrity of anatomical connections within the default network in schizophrenia. Their fMRI data has been co-registered with anatomical scans, and has been put it into DTI space. Figure 8 shows pilot data of the white matter connections through the corpus callosum between the left and the right auditory cortex ROIs as defined by fMRI data. This data will be also used to test robustness of our module, since data was collected on a different scanner (3T Philips).

;* Tractography Comparison Project
We are also working on a [http://www.na-mic.org/Wiki/index.php/SanteFe.Tractography.Conference tractography comparison project]dataset, where we apply stochastic tractography to phantom, as well as test dataset.

===Staffing Plan===

* Sylvain and Yogesh are the DBP resources charged with adapting the tools in the NA-MIC Kit to the DBP needs
* Doug Markant, our NAMIC RA has left the lab, and now Doug Terry, is a new NAMIC RA.
* Julien is our new NAMIC software engineer. He is responsible for improving the STM (Stochastic Tractography Module), and making sure software works with STM compliant datasets. [[Link Progress| Progress]]
* Polina is the algorithm core contact
* Brad is the engineering core contact

===Schedule===

* '''10/2007''' - Optimization of Stochastic Tractography algorythm for 1.5T data.
* '''10/2007''' - Algorythm testing on Santa Fe data set and diffusion phantom.
* '''06/2008''' - Optimization of Stochastic Tractography algorythm for 3T data.
* '''11/2008''' - Slicer 3 module prototype using python.
* '''12/2008''' - Slicer 3 module official release
* '''12/2008''' - Documentation and packaging for dissemination.
* '''12/2008''' - Arcuate Fasciculus results.
* '''01/2009''' - Arcuate Fasciuclus first draft of the paper.
* '''05/2009''' - Distortion correction and nonlinear registration added to the module
* '''05/2009''' - Symposium on tractography, including stochastic methods at World Biological Psychiatry Symposium in Florence, Italy.
* '''05/2009''' - Presentation of Arcuate Fasciculus findings at World Biological Psychiatry Symposium in Florence, Italy.

===Team and Institute===
*PI: Marek Kubicki (kubicki at bwh.harvard.edu)
*DBP2 Investigators: Sylvain Bouix, Yogesh Rathi, Julien de Siebenthal
*NA-MIC Engineering Contact: Brad Davis, Kitware
*NA-MIC Algorithms Contact: Polina Gollard, MIT

===Publications===

''In print''

*[http://www.na-mic.org/pages/Special:Publications?text=Kubicki+AND+Westin+AND+DTI&submit=Search&words=all&title=checked&keywords=checked&authors=checked&abstract=checked&sponsors=checked&searchbytag=checked| NA-MIC Publications Database - Clinical Applications]

*[http://www.na-mic.org/pages/Special:Publications?text=Ngo+AND+Golland&submit=Search&words=all&title=checked&keywords=checked&authors=checked&abstract=checked&sponsors=checked&searchbytag=checked| NA-MIC Publications Database - Algorithms Development]

[[Category: Schizophrenia]] [[Category: Diffusion MRI]] [[Category: Segmentation]]

DBP2:Harvard:Brain Segmentation Roadmap

2009-04-16T17:55:29Z

Dougt: /* Work Accomplished */

Back to [[NA-MIC_Internal_Collaborations|NA-MIC Collaborations]], [[DBP2:Harvard|Harvard DBP 2]]
__NOTOC__
=Stochastic Tractography for VCFS=
== Roadmap ==

The main goal of this project is to develop end-to-end application that would be used to characterize anatomical connectivity abnormalities in the brain of patients with velocardiofacial syndrome (VCFS), and to link this information with deficits in schizophrenia. This page describes the technology roadmap for stochastic tractography, using newly acquired 3T data, NAMIC tools and slicer 3.

== Algorithm ==

[[Image:IC_sto_new.png|thumb|right|200px|<font size=1> Figure 1: Comparison of deterministic and stochastic tractography algorithms</font>]]
; A-Description
* Most tractography methods estimate fibers by tracing the maximum direction of diffusion. A limitation of this approach is that, in practice, several factors introduce uncertainty in the tracking procedure, including, noise, splitting and crossing fibers, head motion and image artifacts. To address this uncertainty, stochastic tractography methods have been developed to quantify the uncertainty associated with estimated fibers (Bjornemo et al., 2002). Method uses a propagation model based on stochastics and regularization, which allows paths originating at one point to branch and return a probability distribution of possible paths. The method utilizes principles of a statistical Monte Carlo method called Sequential Importance Sampling and Resampling (SISR). Based on probability functions, using a sequential importance sampling technique ([http://lmi.bwh.harvard.edu/papers/pdfs/2002/bjornemoMICCAI02.pdf Bjornemo et al., 2002]), one can generate thousands of fibers starting in the same point by sequentially drawing random step directions. This gives a very rich model of the fiber distribution, as contrasted with single fibers produced by conventional tractography methods. Moreover, from a large number of sampled paths, probability maps can be generated, providing better estimates of connectivity between several anatomical locations. A comparison of the algorithms can be seen here. (Figure 1)

[[Image:StochasticPic.PNG|thumb|right|200px|<font size=1>Figure 2: Stochastic Tractography of uncinate fasciculis on anatomical data (left) and cingulum bungle on fMRI scan (right)</font>]]
; B-Possible Applications
* Since diffusion direction uncertainty within the gray matter is quite significant; principal diffusion direction approaches usually do not work for tracking between two gray matter regions. Thus if one requires finding connections between a priori selected anatomical gray matter regions, defined either by anatomical segmentations (in case of using structural ROI data), or functional activations (in case of megring DTI with fMRI), stochastic tractography seems to be the method of choice. Here is an example of this application to anatomical data (Figure 2, left image) and to fMRI data (Figure 2, right image).

* Stochastic Tractography is also comparable, if not better, in defining large white matter fiber bundles, especially those traveling through white matter regions characterized by increased diffusion uncertainty (fiber crossings). Example of such application to internal capsule. (Figure 3)

[[Image:IC-comp-new.png|thumb|right|200px|<font size=1>Figure 3: Streamline vs. Stochastic Tractography of the Internal Capsule</font>]]
; C-References

* [http://lmi.bwh.harvard.edu/papers/pdfs/2002/bjornemoMICCAI02.pdf Björnemo M, Brun A, Kikinis R, Westin CF. Regularized stochastic white matter tractography using diffusion tensor MRI. In Fifth International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI'02). Tokyo, Japan, 2002;435-442.]
* [http://lmi.bwh.harvard.edu/papers/pdfs/2006/frimanTMI06.pdf Friman, O., Farneback, G., Westin CF. A Bayesian Approach for Stochastic White Matter Tractography. IEEE Transactions on Medical Imaging, Vol 25, No. 8, Aug. 2006]

[[Image:stoch_menu.png|thumb|right|200px|<font size=1> Figure 4: Python Stochastic Tractography GUI </font>]]

==Module==
Can be found in: MODULES > PYTHON MODULES > PYTHON STOCHASTIC TRACTOGRAPHY
;Functionality of Python Stochastic Tractography module in Slicer 3.0
* IO:
Module reads files (DWI and ROIs) in nhdr format.
* Smoothing:
One can smooth the DWI data (only Gausian smoothing is supported at this time). We recommend it if the data is noisy.
* Brain Mask:
The Brain mask defines the volume in which the tensor will be computed and the tracts evaluated.
*Diffusion Tensor:
This step calculates the tensor and can output anisotropy indices (FA/Mode/Trace)
*Tractography:
Parameters that need to be adjusted:
:1. The amount of tracts that will be seeded from each voxel (we recommend between 500 and 1000 tracts, depending on the workstation power- 1000 tracts per voxel seeded within the large ROI for high resolution DWI can take a long time to compute).
:2. Maximum tract length (in mm), this can eliminate long, unwanted tracts if the regions for which connection is measured are located close to each other [[Image:stoch_menu2a.png|thumb|right|200px|<font size=1> Figure 5: Python Stochastic Tractography GUI, part 2 </font>]]
:3. Step Size: distance between each re-estimation of tensors, usually between 0.5 and 1 mm
:4. Stopping criteria. This can be used on the top of WM mask, to terminate tracts (in case they really want to travel through the CSF, for example).
*Connectivity Map:
This step creates output probability maps.
:1. binary: each voxel is counted only once if at least one fiber pass through it
:2. cumulative: tracts are summed by voxel independently
:3. weighted: tracts are summed by voxel depending on their length

Then, probability maps can be saved as ROIs, and either used directly, or thresholded (at certain probability, step claimed by few publications to remove noise) in slicer to mask and compute average FA, Mode, Trace for entire connection. Diffusion indices can be also weighted by the probability of connection for each voxel.

:* Module documentation can be found here:
:**[[Media:IJdata.tar.gz|Training Dataset]]
:**[[Media:PNLStochasictNew.ppt|Training Presentation]]
:**[[Media:Helix.zip|Sample Helix Dataset]]

==Work Accomplished==
[[Image:helix.png|thumb|right|200px|<font size=1>Figure 6: Stochastic Tractography on Phantom</font>]]
; A - Optimization and testing of stochastic tractography algorithm :
:* Original methodological paper, as well as our first attempts to use the algorithm (CC+ and matlab scripts) have been done on old "NAMIC" 1.5T LSDI data ([https://portal.nbirn.net:443/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=gotoDirectory&CPq_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FHarvard%2Farchive_morph&up=CPq&JavaScript=enabled Structural MRI] and [https://portal.nbirn.net:443/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=gotoDirectory&up=7li&7li_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FHarvard%2Farchive_diffusion&JavaScript=enabled DTI data]).
:* Tri worked hard on making sure algorithm works on new high resolution 3T data (available here: [https://portal.nbirn.net/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=moveUpDir&gv7_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FPNL%2F3T_strct_dti_fmri%2Fcase01017&up=gv7&JavaScript=enabled|PNL 3T Data]).
:* Tests have been done also on the spiral diffusion phantom, to make sure diffusion directions and scanner coordinates are handled properly by the algorithm (Figure 6).

; B - Clinical Applications

:* Algorithm was used to trace and analyze anterior limb of the internal capsule on 1.5T data. It generated reacher representation of frontal fiber projections, it also turned out to be more sensitive to group differences in white matter integrity that conventional deterministic tractography (see Figure 6).
:* Algorithm was also used to trace smaller white matter fiber tracts, such as Cingulum, Fornix, Uncinate Fasciculus, Arcuate Fasciculus on 3T "Santa Fe" dataset (http://www.na-mic.org/Wiki/index.php/SanteFe.Tractography.Conference)

; C - References

:* [http://www.na-mic.org/Wiki/index.php/Image:IC_posternew.png Shenton, M.E., Ngo, T., Rosenberger, G., Westin, C.F., Levitt, J.J., McCarley, R.W., Kubicki, M. Study of Thalamo-Cortical White Matter Fiber Tract Projections in Schizophrenia Using Diffusion Stochastic Tractography. Poster presented at the 46th Meeting of the American College of Neuropsychopharmacology, Boca Raton, FL, December 2007.]
:* [http://www.na-mic.org/Wiki/index.php/Media:dougt_poster.pdf Terry DP, Rausch AC, Alvarado JL, Melonakos ED, Markant D, Westin CF, Kikinis R, von Siebenthal J, Shenton ME, Kubicki M. White Matter Properties of Emotion Related Connections in Schizophrenia. Poster presented at the 2009 Mysell Poster Day, Dept. of Psychiatry, Harvard Medical School, April 2009]
:* [http://www.na-mic.org/Wiki/index.php/Media:Jorge_poster.pdf Alvarado JL, Terry DP, Markant D, Ngo T, Kikinis R, Westin CF, McCarley RW, Shenton ME, Kubicki M. Study of Language-Related White Matter Tract Connections in Schizophrenia using Diffusion Stochastic Tractography. Poster presented at the 2009 Mysell Poster Day, Dept. of Psychiatry, Harvard Medical School, April 2009]
:* Melonakos ED, Shenton ME, Markant D, Alvarado J, Westin CF, Kubicki M. White Matter Properties of Orbitofrontal Connections in Schizophrenia. Poster being presented at the 64th Meeting of the Society of Biological Psychiatry. Vancouver, BC. May 2009.
:* Kubicki, M. Khan, U., Bobrow, L., O'Donnell, L. Pieper, S. Westin, CF., Shenton, ME. New Methods for Assessing Whole Brain DTI Abnormalities in Schizophrenia. Presentation given at the International Congress of World Psychiatric Association. Florence, Italy. April 2009.
:* Kubicki, M., Markant, D., Ngo, T., Westin, CF., McCarley, RW., Shenton, ME. Study of Language Related White Matter Fiber Tract Projections in Schizophrenia Using Diffusion Stochastic Tractography. Presentation given at the International Congress of World Psychiatric Association. Florence, Italy. April 2009.

== Work in Progress ==
[[Image:Anna.png|thumb|right|200px|<font size=1>Figure 7: Max Plank data showing tracts through the corpus connecting 2 cortical ROIs defined by fMRI activations.</font>]]
; A - Optimization and Testing of stochastic tractography module :
:* Julien is testing now separate components of work flow, including the masks, and impact of their precision on stochastic output, as well as impact of number of seeding points on tractography results.
:* At the same time, we are testing the module on a Max Plank dataset of Anna Rotarska-Jagiela by looking at the connections through the corpus callosum of the left auditory cortex to the right auditory cortex as defined by fMRI activation maps (Figure 7) as well as running this data for the tractography comparison project.
:* We are discussing possibility of adding project specific functionality to module, such as third ROI to guide tractography, nonlinear registration button for merging fMRI and DTI data.

; B - Related Clinical Projects

;* Arcuate Fasciculus Extraction Project
[[Image:STArcuate.jpg|thumb|right|200px|<font size=1>Figure 8: The arcuate fasciculus including seed, midpoint and target ROI's.</font>]]
We have started the project of investigating Arcuate Fasciculus using Stochastic Tractography (Figure 8). This structure is especially important in both VCFS and schizophrenia, as it connects language related areas (Brocka and Wernicke's), and is involved in language processing quite disturbed in schizophrenia patients. It also can not be reliably traced using deterministic tractography.
:Project involves:
:* Whole brain segmentation, and automatic extraction of regions interconnected by Arcuate Fasciculus (Inferior frontal and Superior Temporal Gyri).
:* White matter segmentation, in order to prevent algorithm from traveling through the ventricles, where diffusivity is high.
:* Non-linear registration of labelmaps to the DTI space.
:* Seeding tracts. We have piloted it using 5000 seeds per voxel, however it is quite time consuming running it on even most powerful computers in the lab, so we have experimented with smaller number of seeds per voxel. We tested 1000 seeds, which gave virtually identical results.
:* Extracting path of interest, and calculating FA along the path for group comparison. Presentation of previous results for 7 schizophrenics and 12 control subjects, can be found here: [[Media:NAMIC_AHM_Arcuate.ppt|Progress Report Presentation]].
:* Results presentation and paper submission. Abstract was submitted and accepted for presentation at World Biological Psychiatry Symposium in Venice, Italy. Paper is in preparation.

;* Semantic Network Connectivity Project
We use combination of fMRI and DTI data to define and characterize functional and anatomical connectivity within the semantic processing network in schizophrenia.
:Project involves:
:* fMRI data analysis and identification of functional nodes involved in semantic processing in healthy controls and seubjects with schizophrenia
:* Analysis of functional connectivity (using FSL) between nodes of semantic network
:* Whole brain Voxel Based analysis of DTI data in same population
:* Use of stochastic tractography to identify connections between functional nodes
:* Correlational analysis involving anatomical and functional connectivity data.
:* Results presentation and paper submission. Data was presented at HBM conference in 2007, paper has been submitted to HBM.

;* Study of Default Network
We have started collaboration with Department of Neurophysiology Max Planck Institute in Frankfurt (Anna Rotarska contact person). They have dataset containing DTI and resting state fMRI in schizophrenia, and want to use stochastic tractography to measure integrity of anatomical connections within the default network in schizophrenia. Their fMRI data has been co-registered with anatomical scans, and we are putting it into the DTI space. Once this is done, we will start creating tracts connecting fMRI ROIs. This data will be also used to test robustness of our module, since data was collected on a different scanner (3T Philips).

;* Tractography Comparison Project
We are also working on a [http://www.na-mic.org/Wiki/index.php/SanteFe.Tractography.Conference tractography comparison project]dataset, where we apply stochastic tractography to phantom, as well as test dataset.

===Staffing Plan===

* Sylvain and Yogesh are the DBP resources charged with adapting the tools in the NA-MIC Kit to the DBP needs
* Doug Markant, our NAMIC RA has left the lab, and now Doug Terry, is a new NAMIC RA.
* Julien is our new NAMIC software engineer. He is responsible for improving the STM (Stochastic Tractography Module), and making sure software works with STM compliant datasets. [[Link Progress| Progress]]
* Polina is the algorithm core contact
* Brad is the engineering core contact

===Schedule===

* '''10/2007''' - Optimization of Stochastic Tractography algorythm for 1.5T data.
* '''10/2007''' - Algorythm testing on Santa Fe data set and diffusion phantom.
* '''06/2008''' - Optimization of Stochastic Tractography algorythm for 3T data.
* '''11/2008''' - Slicer 3 module prototype using python.
* '''12/2008''' - Slicer 3 module official release
* '''12/2008''' - Documentation and packaging for dissemination.
* '''12/2008''' - Arcuate Fasciculus results.
* '''01/2009''' - Arcuate Fasciuclus first draft of the paper.
* '''05/2009''' - Distortion correction and nonlinear registration added to the module
* '''05/2009''' - Symposium on tractography, including stochastic methods at World Biological Psychiatry Symposium in Florence, Italy.
* '''05/2009''' - Presentation of Arcuate Fasciculus findings at World Biological Psychiatry Symposium in Florence, Italy.

===Team and Institute===
*PI: Marek Kubicki (kubicki at bwh.harvard.edu)
*DBP2 Investigators: Sylvain Bouix, Yogesh Rathi, Julien de Siebenthal
*NA-MIC Engineering Contact: Brad Davis, Kitware
*NA-MIC Algorithms Contact: Polina Gollard, MIT

===Publications===

''In print''

*[http://www.na-mic.org/pages/Special:Publications?text=Kubicki+AND+Westin+AND+DTI&submit=Search&words=all&title=checked&keywords=checked&authors=checked&abstract=checked&sponsors=checked&searchbytag=checked| NA-MIC Publications Database - Clinical Applications]

*[http://www.na-mic.org/pages/Special:Publications?text=Ngo+AND+Golland&submit=Search&words=all&title=checked&keywords=checked&authors=checked&abstract=checked&sponsors=checked&searchbytag=checked| NA-MIC Publications Database - Algorithms Development]

[[Category: Schizophrenia]] [[Category: Diffusion MRI]] [[Category: Segmentation]]

DBP2:Harvard:Brain Segmentation Roadmap

2009-04-16T17:55:15Z

Dougt: /* Work Accomplished */

Back to [[NA-MIC_Internal_Collaborations|NA-MIC Collaborations]], [[DBP2:Harvard|Harvard DBP 2]]
__NOTOC__
=Stochastic Tractography for VCFS=
== Roadmap ==

The main goal of this project is to develop end-to-end application that would be used to characterize anatomical connectivity abnormalities in the brain of patients with velocardiofacial syndrome (VCFS), and to link this information with deficits in schizophrenia. This page describes the technology roadmap for stochastic tractography, using newly acquired 3T data, NAMIC tools and slicer 3.

== Algorithm ==

[[Image:IC_sto_new.png|thumb|right|200px|<font size=1> Figure 1: Comparison of deterministic and stochastic tractography algorithms</font>]]
; A-Description
* Most tractography methods estimate fibers by tracing the maximum direction of diffusion. A limitation of this approach is that, in practice, several factors introduce uncertainty in the tracking procedure, including, noise, splitting and crossing fibers, head motion and image artifacts. To address this uncertainty, stochastic tractography methods have been developed to quantify the uncertainty associated with estimated fibers (Bjornemo et al., 2002). Method uses a propagation model based on stochastics and regularization, which allows paths originating at one point to branch and return a probability distribution of possible paths. The method utilizes principles of a statistical Monte Carlo method called Sequential Importance Sampling and Resampling (SISR). Based on probability functions, using a sequential importance sampling technique ([http://lmi.bwh.harvard.edu/papers/pdfs/2002/bjornemoMICCAI02.pdf Bjornemo et al., 2002]), one can generate thousands of fibers starting in the same point by sequentially drawing random step directions. This gives a very rich model of the fiber distribution, as contrasted with single fibers produced by conventional tractography methods. Moreover, from a large number of sampled paths, probability maps can be generated, providing better estimates of connectivity between several anatomical locations. A comparison of the algorithms can be seen here. (Figure 1)

[[Image:StochasticPic.PNG|thumb|right|200px|<font size=1>Figure 2: Stochastic Tractography of uncinate fasciculis on anatomical data (left) and cingulum bungle on fMRI scan (right)</font>]]
; B-Possible Applications
* Since diffusion direction uncertainty within the gray matter is quite significant; principal diffusion direction approaches usually do not work for tracking between two gray matter regions. Thus if one requires finding connections between a priori selected anatomical gray matter regions, defined either by anatomical segmentations (in case of using structural ROI data), or functional activations (in case of megring DTI with fMRI), stochastic tractography seems to be the method of choice. Here is an example of this application to anatomical data (Figure 2, left image) and to fMRI data (Figure 2, right image).

* Stochastic Tractography is also comparable, if not better, in defining large white matter fiber bundles, especially those traveling through white matter regions characterized by increased diffusion uncertainty (fiber crossings). Example of such application to internal capsule. (Figure 3)

[[Image:IC-comp-new.png|thumb|right|200px|<font size=1>Figure 3: Streamline vs. Stochastic Tractography of the Internal Capsule</font>]]
; C-References

* [http://lmi.bwh.harvard.edu/papers/pdfs/2002/bjornemoMICCAI02.pdf Björnemo M, Brun A, Kikinis R, Westin CF. Regularized stochastic white matter tractography using diffusion tensor MRI. In Fifth International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI'02). Tokyo, Japan, 2002;435-442.]
* [http://lmi.bwh.harvard.edu/papers/pdfs/2006/frimanTMI06.pdf Friman, O., Farneback, G., Westin CF. A Bayesian Approach for Stochastic White Matter Tractography. IEEE Transactions on Medical Imaging, Vol 25, No. 8, Aug. 2006]

[[Image:stoch_menu.png|thumb|right|200px|<font size=1> Figure 4: Python Stochastic Tractography GUI </font>]]

==Module==
Can be found in: MODULES > PYTHON MODULES > PYTHON STOCHASTIC TRACTOGRAPHY
;Functionality of Python Stochastic Tractography module in Slicer 3.0
* IO:
Module reads files (DWI and ROIs) in nhdr format.
* Smoothing:
One can smooth the DWI data (only Gausian smoothing is supported at this time). We recommend it if the data is noisy.
* Brain Mask:
The Brain mask defines the volume in which the tensor will be computed and the tracts evaluated.
*Diffusion Tensor:
This step calculates the tensor and can output anisotropy indices (FA/Mode/Trace)
*Tractography:
Parameters that need to be adjusted:
:1. The amount of tracts that will be seeded from each voxel (we recommend between 500 and 1000 tracts, depending on the workstation power- 1000 tracts per voxel seeded within the large ROI for high resolution DWI can take a long time to compute).
:2. Maximum tract length (in mm), this can eliminate long, unwanted tracts if the regions for which connection is measured are located close to each other [[Image:stoch_menu2a.png|thumb|right|200px|<font size=1> Figure 5: Python Stochastic Tractography GUI, part 2 </font>]]
:3. Step Size: distance between each re-estimation of tensors, usually between 0.5 and 1 mm
:4. Stopping criteria. This can be used on the top of WM mask, to terminate tracts (in case they really want to travel through the CSF, for example).
*Connectivity Map:
This step creates output probability maps.
:1. binary: each voxel is counted only once if at least one fiber pass through it
:2. cumulative: tracts are summed by voxel independently
:3. weighted: tracts are summed by voxel depending on their length

Then, probability maps can be saved as ROIs, and either used directly, or thresholded (at certain probability, step claimed by few publications to remove noise) in slicer to mask and compute average FA, Mode, Trace for entire connection. Diffusion indices can be also weighted by the probability of connection for each voxel.

:* Module documentation can be found here:
:**[[Media:IJdata.tar.gz|Training Dataset]]
:**[[Media:PNLStochasictNew.ppt|Training Presentation]]
:**[[Media:Helix.zip|Sample Helix Dataset]]

==Work Accomplished==
[[Image:helix.png|thumb|right|200px|<font size=1>Figure 6: Stochastic Tractography on Phantom</font>]]
; A - Optimization and testing of stochastic tractography algorithm :
:* Original methodological paper, as well as our first attempts to use the algorithm (CC+ and matlab scripts) have been done on old "NAMIC" 1.5T LSDI data ([https://portal.nbirn.net:443/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=gotoDirectory&CPq_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FHarvard%2Farchive_morph&up=CPq&JavaScript=enabled Structural MRI] and [https://portal.nbirn.net:443/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=gotoDirectory&up=7li&7li_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FHarvard%2Farchive_diffusion&JavaScript=enabled DTI data]).
:* Tri worked hard on making sure algorithm works on new high resolution 3T data (available here: [https://portal.nbirn.net/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=moveUpDir&gv7_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FPNL%2F3T_strct_dti_fmri%2Fcase01017&up=gv7&JavaScript=enabled|PNL 3T Data]).
:* Tests have been done also on the spiral diffusion phantom, to make sure diffusion directions and scanner coordinates are handled properly by the algorithm (Figure 6).

; B - Clinical Applications

:* Algorithm was used to trace and analyze anterior limb of the internal capsule on 1.5T data. It generated reacher representation of frontal fiber projections, it also turned out to be more sensitive to group differences in white matter integrity that conventional deterministic tractography (see Figure 6).
:* Algorithm was also used to trace smaller white matter fiber tracts, such as Cingulum, Fornix, Uncinate Fasciculus, Arcuate Fasciculus on 3T "Santa Fe" dataset (http://www.na-mic.org/Wiki/index.php/SanteFe.Tractography.Conference)

; C - References

:* [http://www.na-mic.org/Wiki/index.php/Image:IC_posternew.png Shenton, M.E., Ngo, T., Rosenberger, G., Westin, C.F., Levitt, J.J., McCarley, R.W., Kubicki, M. Study of Thalamo-Cortical White Matter Fiber Tract Projections in Schizophrenia Using Diffusion Stochastic Tractography. Poster presented at the 46th Meeting of the American College of Neuropsychopharmacology, Boca Raton, FL, December 2007.]
:* [http://www.na-mic.org/Wiki/index.php/Media:dougt_poster.pdf Terry DP, Rausch AC, Alvarado JL, Melonakos ED, Markant D, Westin CF, Kikinis R, von Siebenthal J, Shenton ME, Kubicki M. White Matter Properties of Emotion Related Connections in Schizophrenia. Poster presented at the 2009 Mysell Poster Day, Dept. of Psychiatry, Harvard Medical School, April 2009]
:* [http://www.na-mic.org/Wiki/index.php/Media:Jorge_poster.pdf Alvarado JL, Terry DP, Markant D, Ngo T, Kikinis R, Westin CF, McCarley RW, Shenton ME, Kubicki M. Study of Language-Related White Matter Tract Connections in Schizophrenia using Diffusion Stochastic Tractography. Poster presented at the 2009 Mysell Poster Day, Dept. of Psychiatry, Harvard Medical School, April 2009]
:* Melonakos ED, Shenton ME, Markant D, Alvarado J, Westin CF, Kubicki M. White Matter Properties of Orbitofrontal Connections in Schizophrenia. Poster being presented at the 64th Meeting of the Society of Biological Psychiatry. Vancouver, BC. May 2009.
:* Kubicki, M. Khan, U., Bobrow, L., O'Donnell, L. Pieper, S. Westin, CF., Shenton, ME. New Methods for Assessing Whole Brain DTI Abnormalities in Schizophrenia. Presentation given at the International Congress of World Psychiatric Association. Florence, Italy. April 2009.
:* Kubicki, M., Markant, D., Ngo, T., Westin, CF., McCarley, RW., Shenton, ME. Study of Language Related White Matter Fiber Tract Projections in Schizophrenia Using Diffusion Stochastic Tractography.

== Work in Progress ==
[[Image:Anna.png|thumb|right|200px|<font size=1>Figure 7: Max Plank data showing tracts through the corpus connecting 2 cortical ROIs defined by fMRI activations.</font>]]
; A - Optimization and Testing of stochastic tractography module :
:* Julien is testing now separate components of work flow, including the masks, and impact of their precision on stochastic output, as well as impact of number of seeding points on tractography results.
:* At the same time, we are testing the module on a Max Plank dataset of Anna Rotarska-Jagiela by looking at the connections through the corpus callosum of the left auditory cortex to the right auditory cortex as defined by fMRI activation maps (Figure 7) as well as running this data for the tractography comparison project.
:* We are discussing possibility of adding project specific functionality to module, such as third ROI to guide tractography, nonlinear registration button for merging fMRI and DTI data.

; B - Related Clinical Projects

;* Arcuate Fasciculus Extraction Project
[[Image:STArcuate.jpg|thumb|right|200px|<font size=1>Figure 8: The arcuate fasciculus including seed, midpoint and target ROI's.</font>]]
We have started the project of investigating Arcuate Fasciculus using Stochastic Tractography (Figure 8). This structure is especially important in both VCFS and schizophrenia, as it connects language related areas (Brocka and Wernicke's), and is involved in language processing quite disturbed in schizophrenia patients. It also can not be reliably traced using deterministic tractography.
:Project involves:
:* Whole brain segmentation, and automatic extraction of regions interconnected by Arcuate Fasciculus (Inferior frontal and Superior Temporal Gyri).
:* White matter segmentation, in order to prevent algorithm from traveling through the ventricles, where diffusivity is high.
:* Non-linear registration of labelmaps to the DTI space.
:* Seeding tracts. We have piloted it using 5000 seeds per voxel, however it is quite time consuming running it on even most powerful computers in the lab, so we have experimented with smaller number of seeds per voxel. We tested 1000 seeds, which gave virtually identical results.
:* Extracting path of interest, and calculating FA along the path for group comparison. Presentation of previous results for 7 schizophrenics and 12 control subjects, can be found here: [[Media:NAMIC_AHM_Arcuate.ppt|Progress Report Presentation]].
:* Results presentation and paper submission. Abstract was submitted and accepted for presentation at World Biological Psychiatry Symposium in Venice, Italy. Paper is in preparation.

;* Semantic Network Connectivity Project
We use combination of fMRI and DTI data to define and characterize functional and anatomical connectivity within the semantic processing network in schizophrenia.
:Project involves:
:* fMRI data analysis and identification of functional nodes involved in semantic processing in healthy controls and seubjects with schizophrenia
:* Analysis of functional connectivity (using FSL) between nodes of semantic network
:* Whole brain Voxel Based analysis of DTI data in same population
:* Use of stochastic tractography to identify connections between functional nodes
:* Correlational analysis involving anatomical and functional connectivity data.
:* Results presentation and paper submission. Data was presented at HBM conference in 2007, paper has been submitted to HBM.

;* Study of Default Network
We have started collaboration with Department of Neurophysiology Max Planck Institute in Frankfurt (Anna Rotarska contact person). They have dataset containing DTI and resting state fMRI in schizophrenia, and want to use stochastic tractography to measure integrity of anatomical connections within the default network in schizophrenia. Their fMRI data has been co-registered with anatomical scans, and we are putting it into the DTI space. Once this is done, we will start creating tracts connecting fMRI ROIs. This data will be also used to test robustness of our module, since data was collected on a different scanner (3T Philips).

;* Tractography Comparison Project
We are also working on a [http://www.na-mic.org/Wiki/index.php/SanteFe.Tractography.Conference tractography comparison project]dataset, where we apply stochastic tractography to phantom, as well as test dataset.

===Staffing Plan===

* Sylvain and Yogesh are the DBP resources charged with adapting the tools in the NA-MIC Kit to the DBP needs
* Doug Markant, our NAMIC RA has left the lab, and now Doug Terry, is a new NAMIC RA.
* Julien is our new NAMIC software engineer. He is responsible for improving the STM (Stochastic Tractography Module), and making sure software works with STM compliant datasets. [[Link Progress| Progress]]
* Polina is the algorithm core contact
* Brad is the engineering core contact

===Schedule===

* '''10/2007''' - Optimization of Stochastic Tractography algorythm for 1.5T data.
* '''10/2007''' - Algorythm testing on Santa Fe data set and diffusion phantom.
* '''06/2008''' - Optimization of Stochastic Tractography algorythm for 3T data.
* '''11/2008''' - Slicer 3 module prototype using python.
* '''12/2008''' - Slicer 3 module official release
* '''12/2008''' - Documentation and packaging for dissemination.
* '''12/2008''' - Arcuate Fasciculus results.
* '''01/2009''' - Arcuate Fasciuclus first draft of the paper.
* '''05/2009''' - Distortion correction and nonlinear registration added to the module
* '''05/2009''' - Symposium on tractography, including stochastic methods at World Biological Psychiatry Symposium in Florence, Italy.
* '''05/2009''' - Presentation of Arcuate Fasciculus findings at World Biological Psychiatry Symposium in Florence, Italy.

===Team and Institute===
*PI: Marek Kubicki (kubicki at bwh.harvard.edu)
*DBP2 Investigators: Sylvain Bouix, Yogesh Rathi, Julien de Siebenthal
*NA-MIC Engineering Contact: Brad Davis, Kitware
*NA-MIC Algorithms Contact: Polina Gollard, MIT

===Publications===

''In print''

*[http://www.na-mic.org/pages/Special:Publications?text=Kubicki+AND+Westin+AND+DTI&submit=Search&words=all&title=checked&keywords=checked&authors=checked&abstract=checked&sponsors=checked&searchbytag=checked| NA-MIC Publications Database - Clinical Applications]

*[http://www.na-mic.org/pages/Special:Publications?text=Ngo+AND+Golland&submit=Search&words=all&title=checked&keywords=checked&authors=checked&abstract=checked&sponsors=checked&searchbytag=checked| NA-MIC Publications Database - Algorithms Development]

[[Category: Schizophrenia]] [[Category: Diffusion MRI]] [[Category: Segmentation]]

DBP2:Harvard:Brain Segmentation Roadmap

2009-04-16T17:49:13Z

Dougt: /* Work in Progress */

Back to [[NA-MIC_Internal_Collaborations|NA-MIC Collaborations]], [[DBP2:Harvard|Harvard DBP 2]]
__NOTOC__
=Stochastic Tractography for VCFS=
== Roadmap ==

The main goal of this project is to develop end-to-end application that would be used to characterize anatomical connectivity abnormalities in the brain of patients with velocardiofacial syndrome (VCFS), and to link this information with deficits in schizophrenia. This page describes the technology roadmap for stochastic tractography, using newly acquired 3T data, NAMIC tools and slicer 3.

== Algorithm ==

[[Image:IC_sto_new.png|thumb|right|200px|<font size=1> Figure 1: Comparison of deterministic and stochastic tractography algorithms</font>]]
; A-Description
* Most tractography methods estimate fibers by tracing the maximum direction of diffusion. A limitation of this approach is that, in practice, several factors introduce uncertainty in the tracking procedure, including, noise, splitting and crossing fibers, head motion and image artifacts. To address this uncertainty, stochastic tractography methods have been developed to quantify the uncertainty associated with estimated fibers (Bjornemo et al., 2002). Method uses a propagation model based on stochastics and regularization, which allows paths originating at one point to branch and return a probability distribution of possible paths. The method utilizes principles of a statistical Monte Carlo method called Sequential Importance Sampling and Resampling (SISR). Based on probability functions, using a sequential importance sampling technique ([http://lmi.bwh.harvard.edu/papers/pdfs/2002/bjornemoMICCAI02.pdf Bjornemo et al., 2002]), one can generate thousands of fibers starting in the same point by sequentially drawing random step directions. This gives a very rich model of the fiber distribution, as contrasted with single fibers produced by conventional tractography methods. Moreover, from a large number of sampled paths, probability maps can be generated, providing better estimates of connectivity between several anatomical locations. A comparison of the algorithms can be seen here. (Figure 1)

[[Image:StochasticPic.PNG|thumb|right|200px|<font size=1>Figure 2: Stochastic Tractography of uncinate fasciculis on anatomical data (left) and cingulum bungle on fMRI scan (right)</font>]]
; B-Possible Applications
* Since diffusion direction uncertainty within the gray matter is quite significant; principal diffusion direction approaches usually do not work for tracking between two gray matter regions. Thus if one requires finding connections between a priori selected anatomical gray matter regions, defined either by anatomical segmentations (in case of using structural ROI data), or functional activations (in case of megring DTI with fMRI), stochastic tractography seems to be the method of choice. Here is an example of this application to anatomical data (Figure 2, left image) and to fMRI data (Figure 2, right image).

* Stochastic Tractography is also comparable, if not better, in defining large white matter fiber bundles, especially those traveling through white matter regions characterized by increased diffusion uncertainty (fiber crossings). Example of such application to internal capsule. (Figure 3)

[[Image:IC-comp-new.png|thumb|right|200px|<font size=1>Figure 3: Streamline vs. Stochastic Tractography of the Internal Capsule</font>]]
; C-References

* [http://lmi.bwh.harvard.edu/papers/pdfs/2002/bjornemoMICCAI02.pdf Björnemo M, Brun A, Kikinis R, Westin CF. Regularized stochastic white matter tractography using diffusion tensor MRI. In Fifth International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI'02). Tokyo, Japan, 2002;435-442.]
* [http://lmi.bwh.harvard.edu/papers/pdfs/2006/frimanTMI06.pdf Friman, O., Farneback, G., Westin CF. A Bayesian Approach for Stochastic White Matter Tractography. IEEE Transactions on Medical Imaging, Vol 25, No. 8, Aug. 2006]

[[Image:stoch_menu.png|thumb|right|200px|<font size=1> Figure 4: Python Stochastic Tractography GUI </font>]]

==Module==
Can be found in: MODULES > PYTHON MODULES > PYTHON STOCHASTIC TRACTOGRAPHY
;Functionality of Python Stochastic Tractography module in Slicer 3.0
* IO:
Module reads files (DWI and ROIs) in nhdr format.
* Smoothing:
One can smooth the DWI data (only Gausian smoothing is supported at this time). We recommend it if the data is noisy.
* Brain Mask:
The Brain mask defines the volume in which the tensor will be computed and the tracts evaluated.
*Diffusion Tensor:
This step calculates the tensor and can output anisotropy indices (FA/Mode/Trace)
*Tractography:
Parameters that need to be adjusted:
:1. The amount of tracts that will be seeded from each voxel (we recommend between 500 and 1000 tracts, depending on the workstation power- 1000 tracts per voxel seeded within the large ROI for high resolution DWI can take a long time to compute).
:2. Maximum tract length (in mm), this can eliminate long, unwanted tracts if the regions for which connection is measured are located close to each other [[Image:stoch_menu2a.png|thumb|right|200px|<font size=1> Figure 5: Python Stochastic Tractography GUI, part 2 </font>]]
:3. Step Size: distance between each re-estimation of tensors, usually between 0.5 and 1 mm
:4. Stopping criteria. This can be used on the top of WM mask, to terminate tracts (in case they really want to travel through the CSF, for example).
*Connectivity Map:
This step creates output probability maps.
:1. binary: each voxel is counted only once if at least one fiber pass through it
:2. cumulative: tracts are summed by voxel independently
:3. weighted: tracts are summed by voxel depending on their length

Then, probability maps can be saved as ROIs, and either used directly, or thresholded (at certain probability, step claimed by few publications to remove noise) in slicer to mask and compute average FA, Mode, Trace for entire connection. Diffusion indices can be also weighted by the probability of connection for each voxel.

:* Module documentation can be found here:
:**[[Media:IJdata.tar.gz|Training Dataset]]
:**[[Media:PNLStochasictNew.ppt|Training Presentation]]
:**[[Media:Helix.zip|Sample Helix Dataset]]

==Work Accomplished==
[[Image:helix.png|thumb|right|200px|<font size=1>Figure 6: Stochastic Tractography on Phantom</font>]]
; A - Optimization and testing of stochastic tractography algorithm :
:* Original methodological paper, as well as our first attempts to use the algorithm (CC+ and matlab scripts) have been done on old "NAMIC" 1.5T LSDI data ([https://portal.nbirn.net:443/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=gotoDirectory&CPq_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FHarvard%2Farchive_morph&up=CPq&JavaScript=enabled Structural MRI] and [https://portal.nbirn.net:443/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=gotoDirectory&up=7li&7li_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FHarvard%2Farchive_diffusion&JavaScript=enabled DTI data]).
:* Tri worked hard on making sure algorithm works on new high resolution 3T data (available here: [https://portal.nbirn.net/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=moveUpDir&gv7_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FPNL%2F3T_strct_dti_fmri%2Fcase01017&up=gv7&JavaScript=enabled|PNL 3T Data]).
:* Tests have been done also on the spiral diffusion phantom, to make sure diffusion directions and scanner coordinates are handled properly by the algorithm (Figure 6).

; B - Clinical Applications

:* Algorithm was used to trace and analyze anterior limb of the internal capsule on 1.5T data. It generated reacher representation of frontal fiber projections, it also turned out to be more sensitive to group differences in white matter integrity that conventional deterministic tractography (see Figure 6).
:* Algorithm was also used to trace smaller white matter fiber tracts, such as Cingulum, Fornix, Uncinate Fasciculus, Arcuate Fasciculus on 3T "Santa Fe" dataset (http://www.na-mic.org/Wiki/index.php/SanteFe.Tractography.Conference)

; C - References

:* [http://www.na-mic.org/Wiki/index.php/Image:IC_posternew.png Shenton, M.E., Ngo, T., Rosenberger, G., Westin, C.F., Levitt, J.J., McCarley, R.W., Kubicki, M. Study of Thalamo-Cortical White Matter Fiber Tract Projections in Schizophrenia Using Diffusion Stochastic Tractography. Poster presented at the 46th Meeting of the American College of Neuropsychopharmacology, Boca Raton, FL, December 2007.]
:* [http://www.na-mic.org/Wiki/index.php/Media:dougt_poster.pdf Terry DP, Rausch AC, Alvarado JL, Melonakos ED, Markant D, Westin CF, Kikinis R, von Siebenthal J, Shenton ME, Kubicki M. White Matter Properties of Emotion Related Connections in Schizophrenia. Poster presented at the 2009 Mysell Poster Day, Dept. of Psychiatry, Harvard Medical School, April 2009]
:* [http://www.na-mic.org/Wiki/index.php/Media:Jorge_poster.pdf Alvarado JL, Terry DP, Markant D, Ngo T, Kikinis R, Westin CF, McCarley RW, Shenton ME, Kubicki M. Study of Language-Related White Matter Tract Connections in Schizophrenia using Diffusion Stochastic Tractography. Poster presented at the 2009 Mysell Poster Day, Dept. of Psychiatry, Harvard Medical School, April 2009]
:* Melonakos ED, Shenton ME, Markant D, Alvarado J, Westin CF, Kubicki M. White Matter Properties of Orbitofrontal Connections in Schizophrenia. Poster being presented at the 64th Meeting of the Society of Biological Psychiatry. Vancouver, BC. May 2009.
:* Kubicki, M. Advances in DTI and Its Applications to Schizophrenia. Presentation given at the International Congress of World Psychiatric Association meeting, Florence, Italy. April 2009.
:* Kubicki, M. New Methods for Assessing Whole Brain DTI Abnormalities in Schizophrenia. Presentation given at the International Congress of World Psychiatric Association meeting, Florence, Italy. April 2009.

== Work in Progress ==
[[Image:Anna.png|thumb|right|200px|<font size=1>Figure 7: Max Plank data showing tracts through the corpus connecting 2 cortical ROIs defined by fMRI activations.</font>]]
; A - Optimization and Testing of stochastic tractography module :
:* Julien is testing now separate components of work flow, including the masks, and impact of their precision on stochastic output, as well as impact of number of seeding points on tractography results.
:* At the same time, we are testing the module on a Max Plank dataset of Anna Rotarska-Jagiela by looking at the connections through the corpus callosum of the left auditory cortex to the right auditory cortex as defined by fMRI activation maps (Figure 7) as well as running this data for the tractography comparison project.
:* We are discussing possibility of adding project specific functionality to module, such as third ROI to guide tractography, nonlinear registration button for merging fMRI and DTI data.

; B - Related Clinical Projects

;* Arcuate Fasciculus Extraction Project
[[Image:STArcuate.jpg|thumb|right|200px|<font size=1>Figure 8: The arcuate fasciculus including seed, midpoint and target ROI's.</font>]]
We have started the project of investigating Arcuate Fasciculus using Stochastic Tractography (Figure 8). This structure is especially important in both VCFS and schizophrenia, as it connects language related areas (Brocka and Wernicke's), and is involved in language processing quite disturbed in schizophrenia patients. It also can not be reliably traced using deterministic tractography.
:Project involves:
:* Whole brain segmentation, and automatic extraction of regions interconnected by Arcuate Fasciculus (Inferior frontal and Superior Temporal Gyri).
:* White matter segmentation, in order to prevent algorithm from traveling through the ventricles, where diffusivity is high.
:* Non-linear registration of labelmaps to the DTI space.
:* Seeding tracts. We have piloted it using 5000 seeds per voxel, however it is quite time consuming running it on even most powerful computers in the lab, so we have experimented with smaller number of seeds per voxel. We tested 1000 seeds, which gave virtually identical results.
:* Extracting path of interest, and calculating FA along the path for group comparison. Presentation of previous results for 7 schizophrenics and 12 control subjects, can be found here: [[Media:NAMIC_AHM_Arcuate.ppt|Progress Report Presentation]].
:* Results presentation and paper submission. Abstract was submitted and accepted for presentation at World Biological Psychiatry Symposium in Venice, Italy. Paper is in preparation.

;* Semantic Network Connectivity Project
We use combination of fMRI and DTI data to define and characterize functional and anatomical connectivity within the semantic processing network in schizophrenia.
:Project involves:
:* fMRI data analysis and identification of functional nodes involved in semantic processing in healthy controls and seubjects with schizophrenia
:* Analysis of functional connectivity (using FSL) between nodes of semantic network
:* Whole brain Voxel Based analysis of DTI data in same population
:* Use of stochastic tractography to identify connections between functional nodes
:* Correlational analysis involving anatomical and functional connectivity data.
:* Results presentation and paper submission. Data was presented at HBM conference in 2007, paper has been submitted to HBM.

;* Study of Default Network
We have started collaboration with Department of Neurophysiology Max Planck Institute in Frankfurt (Anna Rotarska contact person). They have dataset containing DTI and resting state fMRI in schizophrenia, and want to use stochastic tractography to measure integrity of anatomical connections within the default network in schizophrenia. Their fMRI data has been co-registered with anatomical scans, and we are putting it into the DTI space. Once this is done, we will start creating tracts connecting fMRI ROIs. This data will be also used to test robustness of our module, since data was collected on a different scanner (3T Philips).

;* Tractography Comparison Project
We are also working on a [http://www.na-mic.org/Wiki/index.php/SanteFe.Tractography.Conference tractography comparison project]dataset, where we apply stochastic tractography to phantom, as well as test dataset.

===Staffing Plan===

* Sylvain and Yogesh are the DBP resources charged with adapting the tools in the NA-MIC Kit to the DBP needs
* Doug Markant, our NAMIC RA has left the lab, and now Doug Terry, is a new NAMIC RA.
* Julien is our new NAMIC software engineer. He is responsible for improving the STM (Stochastic Tractography Module), and making sure software works with STM compliant datasets. [[Link Progress| Progress]]
* Polina is the algorithm core contact
* Brad is the engineering core contact

===Schedule===

* '''10/2007''' - Optimization of Stochastic Tractography algorythm for 1.5T data.
* '''10/2007''' - Algorythm testing on Santa Fe data set and diffusion phantom.
* '''06/2008''' - Optimization of Stochastic Tractography algorythm for 3T data.
* '''11/2008''' - Slicer 3 module prototype using python.
* '''12/2008''' - Slicer 3 module official release
* '''12/2008''' - Documentation and packaging for dissemination.
* '''12/2008''' - Arcuate Fasciculus results.
* '''01/2009''' - Arcuate Fasciuclus first draft of the paper.
* '''05/2009''' - Distortion correction and nonlinear registration added to the module
* '''05/2009''' - Symposium on tractography, including stochastic methods at World Biological Psychiatry Symposium in Florence, Italy.
* '''05/2009''' - Presentation of Arcuate Fasciculus findings at World Biological Psychiatry Symposium in Florence, Italy.

===Team and Institute===
*PI: Marek Kubicki (kubicki at bwh.harvard.edu)
*DBP2 Investigators: Sylvain Bouix, Yogesh Rathi, Julien de Siebenthal
*NA-MIC Engineering Contact: Brad Davis, Kitware
*NA-MIC Algorithms Contact: Polina Gollard, MIT

===Publications===

''In print''

*[http://www.na-mic.org/pages/Special:Publications?text=Kubicki+AND+Westin+AND+DTI&submit=Search&words=all&title=checked&keywords=checked&authors=checked&abstract=checked&sponsors=checked&searchbytag=checked| NA-MIC Publications Database - Clinical Applications]

*[http://www.na-mic.org/pages/Special:Publications?text=Ngo+AND+Golland&submit=Search&words=all&title=checked&keywords=checked&authors=checked&abstract=checked&sponsors=checked&searchbytag=checked| NA-MIC Publications Database - Algorithms Development]

[[Category: Schizophrenia]] [[Category: Diffusion MRI]] [[Category: Segmentation]]

File:Anna.png

2009-04-16T17:28:57Z

Dougt:

DBP2:Harvard:Brain Segmentation Roadmap

2009-04-16T15:35:21Z

Dougt:

Back to [[NA-MIC_Internal_Collaborations|NA-MIC Collaborations]], [[DBP2:Harvard|Harvard DBP 2]]
__NOTOC__
=Stochastic Tractography for VCFS=
== Roadmap ==

The main goal of this project is to develop end-to-end application that would be used to characterize anatomical connectivity abnormalities in the brain of patients with velocardiofacial syndrome (VCFS), and to link this information with deficits in schizophrenia. This page describes the technology roadmap for stochastic tractography, using newly acquired 3T data, NAMIC tools and slicer 3.

== Algorithm ==

[[Image:IC_sto_new.png|thumb|right|200px|<font size=1> Figure 1: Comparison of deterministic and stochastic tractography algorithms</font>]]
; A-Description
* Most tractography methods estimate fibers by tracing the maximum direction of diffusion. A limitation of this approach is that, in practice, several factors introduce uncertainty in the tracking procedure, including, noise, splitting and crossing fibers, head motion and image artifacts. To address this uncertainty, stochastic tractography methods have been developed to quantify the uncertainty associated with estimated fibers (Bjornemo et al., 2002). Method uses a propagation model based on stochastics and regularization, which allows paths originating at one point to branch and return a probability distribution of possible paths. The method utilizes principles of a statistical Monte Carlo method called Sequential Importance Sampling and Resampling (SISR). Based on probability functions, using a sequential importance sampling technique ([http://lmi.bwh.harvard.edu/papers/pdfs/2002/bjornemoMICCAI02.pdf Bjornemo et al., 2002]), one can generate thousands of fibers starting in the same point by sequentially drawing random step directions. This gives a very rich model of the fiber distribution, as contrasted with single fibers produced by conventional tractography methods. Moreover, from a large number of sampled paths, probability maps can be generated, providing better estimates of connectivity between several anatomical locations. A comparison of the algorithms can be seen here. (Figure 1)

[[Image:StochasticPic.PNG|thumb|right|200px|<font size=1>Figure 2: Stochastic Tractography of uncinate fasciculis on anatomical data (left) and cingulum bungle on fMRI scan (right)</font>]]
; B-Possible Applications
* Since diffusion direction uncertainty within the gray matter is quite significant; principal diffusion direction approaches usually do not work for tracking between two gray matter regions. Thus if one requires finding connections between a priori selected anatomical gray matter regions, defined either by anatomical segmentations (in case of using structural ROI data), or functional activations (in case of megring DTI with fMRI), stochastic tractography seems to be the method of choice. Here is an example of this application to anatomical data (Figure 2, left image) and to fMRI data (Figure 2, right image).

* Stochastic Tractography is also comparable, if not better, in defining large white matter fiber bundles, especially those traveling through white matter regions characterized by increased diffusion uncertainty (fiber crossings). Example of such application to internal capsule. (Figure 3)

[[Image:IC-comp-new.png|thumb|right|200px|<font size=1>Figure 3: Streamline vs. Stochastic Tractography of the Internal Capsule</font>]]
; C-References

* [http://lmi.bwh.harvard.edu/papers/pdfs/2002/bjornemoMICCAI02.pdf Björnemo M, Brun A, Kikinis R, Westin CF. Regularized stochastic white matter tractography using diffusion tensor MRI. In Fifth International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI'02). Tokyo, Japan, 2002;435-442.]
* [http://lmi.bwh.harvard.edu/papers/pdfs/2006/frimanTMI06.pdf Friman, O., Farneback, G., Westin CF. A Bayesian Approach for Stochastic White Matter Tractography. IEEE Transactions on Medical Imaging, Vol 25, No. 8, Aug. 2006]

[[Image:stoch_menu.png|thumb|right|200px|<font size=1> Figure 4: Python Stochastic Tractography GUI </font>]]

==Module==
Can be found in: MODULES > PYTHON MODULES > PYTHON STOCHASTIC TRACTOGRAPHY
;Functionality of Python Stochastic Tractography module in Slicer 3.0
* IO:
Module reads files (DWI and ROIs) in nhdr format.
* Smoothing:
One can smooth the DWI data (only Gausian smoothing is supported at this time). We recommend it if the data is noisy.
* Brain Mask:
The Brain mask defines the volume in which the tensor will be computed and the tracts evaluated.
*Diffusion Tensor:
This step calculates the tensor and can output anisotropy indices (FA/Mode/Trace)
*Tractography:
Parameters that need to be adjusted:
:1. The amount of tracts that will be seeded from each voxel (we recommend between 500 and 1000 tracts, depending on the workstation power- 1000 tracts per voxel seeded within the large ROI for high resolution DWI can take a long time to compute).
:2. Maximum tract length (in mm), this can eliminate long, unwanted tracts if the regions for which connection is measured are located close to each other [[Image:stoch_menu2a.png|thumb|right|200px|<font size=1> Figure 5: Python Stochastic Tractography GUI, part 2 </font>]]
:3. Step Size: distance between each re-estimation of tensors, usually between 0.5 and 1 mm
:4. Stopping criteria. This can be used on the top of WM mask, to terminate tracts (in case they really want to travel through the CSF, for example).
*Connectivity Map:
This step creates output probability maps.
:1. binary: each voxel is counted only once if at least one fiber pass through it
:2. cumulative: tracts are summed by voxel independently
:3. weighted: tracts are summed by voxel depending on their length

Then, probability maps can be saved as ROIs, and either used directly, or thresholded (at certain probability, step claimed by few publications to remove noise) in slicer to mask and compute average FA, Mode, Trace for entire connection. Diffusion indices can be also weighted by the probability of connection for each voxel.

:* Module documentation can be found here:
:**[[Media:IJdata.tar.gz|Training Dataset]]
:**[[Media:PNLStochasictNew.ppt|Training Presentation]]
:**[[Media:Helix.zip|Sample Helix Dataset]]

==Work Accomplished==
[[Image:helix.png|thumb|right|200px|<font size=1>Figure 6: Stochastic Tractography on Phantom</font>]]
; A - Optimization and testing of stochastic tractography algorithm :
:* Original methodological paper, as well as our first attempts to use the algorithm (CC+ and matlab scripts) have been done on old "NAMIC" 1.5T LSDI data ([https://portal.nbirn.net:443/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=gotoDirectory&CPq_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FHarvard%2Farchive_morph&up=CPq&JavaScript=enabled Structural MRI] and [https://portal.nbirn.net:443/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=gotoDirectory&up=7li&7li_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FHarvard%2Farchive_diffusion&JavaScript=enabled DTI data]).
:* Tri worked hard on making sure algorithm works on new high resolution 3T data (available here: [https://portal.nbirn.net/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=moveUpDir&gv7_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FPNL%2F3T_strct_dti_fmri%2Fcase01017&up=gv7&JavaScript=enabled|PNL 3T Data]).
:* Tests have been done also on the spiral diffusion phantom, to make sure diffusion directions and scanner coordinates are handled properly by the algorithm (Figure 6).

; B - Clinical Applications

:* Algorithm was used to trace and analyze anterior limb of the internal capsule on 1.5T data. It generated reacher representation of frontal fiber projections, it also turned out to be more sensitive to group differences in white matter integrity that conventional deterministic tractography (see Figure 6).
:* Algorithm was also used to trace smaller white matter fiber tracts, such as Cingulum, Fornix, Uncinate Fasciculus, Arcuate Fasciculus on 3T "Santa Fe" dataset (http://www.na-mic.org/Wiki/index.php/SanteFe.Tractography.Conference)

; C - References

:* [http://www.na-mic.org/Wiki/index.php/Image:IC_posternew.png Shenton, M.E., Ngo, T., Rosenberger, G., Westin, C.F., Levitt, J.J., McCarley, R.W., Kubicki, M. Study of Thalamo-Cortical White Matter Fiber Tract Projections in Schizophrenia Using Diffusion Stochastic Tractography. Poster presented at the 46th Meeting of the American College of Neuropsychopharmacology, Boca Raton, FL, December 2007.]
:* [http://www.na-mic.org/Wiki/index.php/Media:dougt_poster.pdf Terry DP, Rausch AC, Alvarado JL, Melonakos ED, Markant D, Westin CF, Kikinis R, von Siebenthal J, Shenton ME, Kubicki M. White Matter Properties of Emotion Related Connections in Schizophrenia. Poster presented at the 2009 Mysell Poster Day, Dept. of Psychiatry, Harvard Medical School, April 2009]
:* [http://www.na-mic.org/Wiki/index.php/Media:Jorge_poster.pdf Alvarado JL, Terry DP, Markant D, Ngo T, Kikinis R, Westin CF, McCarley RW, Shenton ME, Kubicki M. Study of Language-Related White Matter Tract Connections in Schizophrenia using Diffusion Stochastic Tractography. Poster presented at the 2009 Mysell Poster Day, Dept. of Psychiatry, Harvard Medical School, April 2009]
:* Melonakos ED, Shenton ME, Markant D, Alvarado J, Westin CF, Kubicki M. White Matter Properties of Orbitofrontal Connections in Schizophrenia. Poster being presented at the 64th Meeting of the Society of Biological Psychiatry. Vancouver, BC. May 2009.
:* Kubicki, M. Advances in DTI and Its Applications to Schizophrenia. Presentation given at the International Congress of World Psychiatric Association meeting, Florence, Italy. April 2009.
:* Kubicki, M. New Methods for Assessing Whole Brain DTI Abnormalities in Schizophrenia. Presentation given at the International Congress of World Psychiatric Association meeting, Florence, Italy. April 2009.

== Work in Progress ==

; A - Optimization and Testing of stochastic tractography module :
:* Julien is testing now separate components of work flow, including the masks, and impact of their precision on stochastic output, as well as impact of number of seeding points on tractography results.
:* At the same time, we are testing the module on Max Plank dataset, as well as running it on tractography comparison project dataset.
:* We are discussing possibility of adding project specific functionality to module, such as third ROI to guide tractography, nonlinear registration button for merging fMRI and DTI data.

; B - Related Clinical Projects

;* Arcuate Fasciculus Extraction Project
[[Image:STArcuate.jpg|thumb|right|200px|<font size=1>Figure 7: The arcuate fasciculus including seed, midpoint and target ROI's.</font>]]
We have started the project of investigating Arcuate Fasciculus using Stochastic Tractography (Figure 7). This structure is especially important in both VCFS and schizophrenia, as it connects language related areas (Brocka and Wernicke's), and is involved in language processing quite disturbed in schizophrenia patients. It also can not be reliably traced using deterministic tractography.
:Project involves:
:* Whole brain segmentation, and automatic extraction of regions interconnected by Arcuate Fasciculus (Inferior frontal and Superior Temporal Gyri).
:* White matter segmentation, in order to prevent algorithm from traveling through the ventricles, where diffusivity is high.
:* Non-linear registration of labelmaps to the DTI space.
:* Seeding tracts. We have piloted it using 5000 seeds per voxel, however it is quite time consuming running it on even most powerful computers in the lab, so we have experimented with smaller number of seeds per voxel. We tested 1000 seeds, which gave virtually identical results.
:* Extracting path of interest, and calculating FA along the path for group comparison. Presentation of previous results for 7 schizophrenics and 12 control subjects, can be found here: [[Media:NAMIC_AHM_Arcuate.ppt|Progress Report Presentation]].
:* Results presentation and paper submission. Abstract was submitted and accepted for presentation at World Biological Psychiatry Symposium in Venice, Italy. Paper is in preparation.

;* Semantic Network Connectivity Project
We use combination of fMRI and DTI data to define and characterize functional and anatomical connectivity within the semantic processing network in schizophrenia.
:Project involves:
:* fMRI data analysis and identification of functional nodes involved in semantic processing in healthy controls and seubjects with schizophrenia
:* Analysis of functional connectivity (using FSL) between nodes of semantic network
:* Whole brain Voxel Based analysis of DTI data in same population
:* Use of stochastic tractography to identify connections between functional nodes
:* Correlational analysis involving anatomical and functional connectivity data.
:* Results presentation and paper submission. Data was presented at HBM conference in 2007, paper has been submitted to HBM.

;* Study of Default Network
We have started collaboration with Department of Neurophysiology Max Planck Institute in Frankfurt (Anna Rotarska contact person). They have dataset containing DTI and resting state fMRI in schizophrenia, and want to use stochastic tractography to measure integrity of anatomical connections within the default network in schizophrenia. Their fMRI data has been co-registered with anatomical scans, and we are putting it into the DTI space. Once this is done, we will start creating tracts connecting fMRI ROIs. This data will be also used to test robustness of our module, since data was collected on a different scanner (3T Philips).

;* Tractography Comparison Project
We are also working on a [http://www.na-mic.org/Wiki/index.php/SanteFe.Tractography.Conference tractography comparison project]dataset, where we apply stochastic tractography to phantom, as well as test dataset.

===Staffing Plan===

* Sylvain and Yogesh are the DBP resources charged with adapting the tools in the NA-MIC Kit to the DBP needs
* Doug Markant, our NAMIC RA has left the lab, and now Doug Terry, is a new NAMIC RA.
* Julien is our new NAMIC software engineer. He is responsible for improving the STM (Stochastic Tractography Module), and making sure software works with STM compliant datasets. [[Link Progress| Progress]]
* Polina is the algorithm core contact
* Brad is the engineering core contact

===Schedule===

* '''10/2007''' - Optimization of Stochastic Tractography algorythm for 1.5T data.
* '''10/2007''' - Algorythm testing on Santa Fe data set and diffusion phantom.
* '''06/2008''' - Optimization of Stochastic Tractography algorythm for 3T data.
* '''11/2008''' - Slicer 3 module prototype using python.
* '''12/2008''' - Slicer 3 module official release
* '''12/2008''' - Documentation and packaging for dissemination.
* '''12/2008''' - Arcuate Fasciculus results.
* '''01/2009''' - Arcuate Fasciuclus first draft of the paper.
* '''05/2009''' - Distortion correction and nonlinear registration added to the module
* '''05/2009''' - Symposium on tractography, including stochastic methods at World Biological Psychiatry Symposium in Florence, Italy.
* '''05/2009''' - Presentation of Arcuate Fasciculus findings at World Biological Psychiatry Symposium in Florence, Italy.

===Team and Institute===
*PI: Marek Kubicki (kubicki at bwh.harvard.edu)
*DBP2 Investigators: Sylvain Bouix, Yogesh Rathi, Julien de Siebenthal
*NA-MIC Engineering Contact: Brad Davis, Kitware
*NA-MIC Algorithms Contact: Polina Gollard, MIT

===Publications===

''In print''

*[http://www.na-mic.org/pages/Special:Publications?text=Kubicki+AND+Westin+AND+DTI&submit=Search&words=all&title=checked&keywords=checked&authors=checked&abstract=checked&sponsors=checked&searchbytag=checked| NA-MIC Publications Database - Clinical Applications]

*[http://www.na-mic.org/pages/Special:Publications?text=Ngo+AND+Golland&submit=Search&words=all&title=checked&keywords=checked&authors=checked&abstract=checked&sponsors=checked&searchbytag=checked| NA-MIC Publications Database - Algorithms Development]

[[Category: Schizophrenia]] [[Category: Diffusion MRI]] [[Category: Segmentation]]

DBP2:Harvard:Brain Segmentation Roadmap

2009-04-16T14:56:00Z

Dougt:

Back to [[NA-MIC_Internal_Collaborations|NA-MIC Collaborations]], [[DBP2:Harvard|Harvard DBP 2]]
__NOTOC__
=Stochastic Tractography for VCFS=
== Roadmap ==

The main goal of this project is to develop end-to-end application that would be used to characterize anatomical connectivity abnormalities in the brain of patients with velocardiofacial syndrome (VCFS), and to link this information with deficits in schizophrenia. This page describes the technology roadmap for stochastic tractography, using newly acquired 3T data, NAMIC tools and slicer 3.

== Algorithm ==

[[Image:IC_sto_new.png|thumb|right|200px|<font size=1> Figure 1: Comparison of deterministic and stochastic tractography algorithms</font>]]
; A-Description
* Most tractography methods estimate fibers by tracing the maximum direction of diffusion. A limitation of this approach is that, in practice, several factors introduce uncertainty in the tracking procedure, including, noise, splitting and crossing fibers, head motion and image artifacts. To address this uncertainty, stochastic tractography methods have been developed to quantify the uncertainty associated with estimated fibers (Bjornemo et al., 2002). Method uses a propagation model based on stochastics and regularization, which allows paths originating at one point to branch and return a probability distribution of possible paths. The method utilizes principles of a statistical Monte Carlo method called Sequential Importance Sampling and Resampling (SISR). Based on probability functions, using a sequential importance sampling technique ([http://lmi.bwh.harvard.edu/papers/pdfs/2002/bjornemoMICCAI02.pdf Bjornemo et al., 2002]), one can generate thousands of fibers starting in the same point by sequentially drawing random step directions. This gives a very rich model of the fiber distribution, as contrasted with single fibers produced by conventional tractography methods. Moreover, from a large number of sampled paths, probability maps can be generated, providing better estimates of connectivity between several anatomical locations. A comparison of the algorithms can be seen here. (Figure 1)

[[Image:StochasticPic.PNG|thumb|right|200px|<font size=1>Figure 2: Stochastic Tractography of uncinate fasciculis on anatomical data (left) and cingulum bungle on fMRI scan (right)</font>]]
; B-Possible Applications
* Since diffusion direction uncertainty within the gray matter is quite significant; principal diffusion direction approaches usually do not work for tracking between two gray matter regions. Thus if one requires finding connections between a priori selected anatomical gray matter regions, defined either by anatomical segmentations (in case of using structural ROI data), or functional activations (in case of megring DTI with fMRI), stochastic tractography seems to be the method of choice. Here is an example of this application to anatomical data (Figure 2, left image) and to fMRI data (Figure 2, right image).

* Stochastic Tractography is also comparable, if not better, in defining large white matter fiber bundles, especially those traveling through white matter regions characterized by increased diffusion uncertainty (fiber crossings). Example of such application to internal capsule. (Figure 3)

[[Image:IC-comp-new.png|thumb|right|200px|<font size=1>Figure 3: Streamline vs. Stochastic Tractography of the Internal Capsule</font>]]
; C-References

* [http://lmi.bwh.harvard.edu/papers/pdfs/2002/bjornemoMICCAI02.pdf Björnemo M, Brun A, Kikinis R, Westin CF. Regularized stochastic white matter tractography using diffusion tensor MRI. In Fifth International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI'02). Tokyo, Japan, 2002;435-442.]
* [http://lmi.bwh.harvard.edu/papers/pdfs/2006/frimanTMI06.pdf Friman, O., Farneback, G., Westin CF. A Bayesian Approach for Stochastic White Matter Tractography. IEEE Transactions on Medical Imaging, Vol 25, No. 8, Aug. 2006]

[[Image:stoch_menu.png|thumb|right|200px|<font size=1> Figure 4: Python Stochastic Tractography GUI </font>]]

==Module==
Can be found in: MODULES > PYTHON MODULES > PYTHON STOCHASTIC TRACTOGRAPHY
;Functionality of Python Stochastic Tractography module in Slicer 3.0
* IO:
Module reads files (DWI and ROIs) in nhdr format.
* Smoothing:
One can smooth the DWI data (only Gausian smoothing is supported at this time). We recommend it if the data is noisy.
* Brain Mask:
The Brain mask defines the volume in which the tensor will be computed and the tracts evaluated.
*Diffusion Tensor:
This step calculates the tensor and can output anisotropy indices (FA/Mode/Trace)
*Tractography:
Parameters that need to be adjusted:
:1. The amount of tracts that will be seeded from each voxel (we recommend between 500 and 1000 tracts, depending on the workstation power- 1000 tracts per voxel seeded within the large ROI for high resolution DWI can take a long time to compute).
:2. Maximum tract length (in mm), this can eliminate long, unwanted tracts if the regions for which connection is measured are located close to each other [[Image:stoch_menu2a.png|thumb|right|200px|<font size=1> Figure 5: Python Stochastic Tractography GUI, part 2 </font>]]
:3. Step Size: distance between each re-estimation of tensors, usually between 0.5 and 1 mm
:4. Stopping criteria. This can be used on the top of WM mask, to terminate tracts (in case they really want to travel through the CSF, for example).
*Connectivity Map:
This step creates output probability maps.
:1. binary: each voxel is counted only once if at least one fiber pass through it
:2. cumulative: tracts are summed by voxel independently
:3. weighted: tracts are summed by voxel depending on their length

Then, probability maps can be saved as ROIs, and either used directly, or thresholded (at certain probability, step claimed by few publications to remove noise) in slicer to mask and compute average FA, Mode, Trace for entire connection. Diffusion indices can be also weighted by the probability of connection for each voxel.

:* Module documentation can be found here:
:**[[Media:IJdata.tar.gz|Training Dataset]]
:**[[Media:PNLStochasictNew.ppt|Training Presentation]]
:**[[Media:Helix.zip|Sample Helix Dataset]]

==Work Accomplished==
[[Image:helix.png|thumb|right|200px|<font size=1>Figure 6: Stochastic Tractography on Phantom</font>]]
; A - Optimization and testing of stochastic tractography algorithm :
:* Original methodological paper, as well as our first attempts to use the algorithm (CC+ and matlab scripts) have been done on old "NAMIC" 1.5T LSDI data ([https://portal.nbirn.net:443/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=gotoDirectory&CPq_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FHarvard%2Farchive_morph&up=CPq&JavaScript=enabled Structural MRI] and [https://portal.nbirn.net:443/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=gotoDirectory&up=7li&7li_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FHarvard%2Farchive_diffusion&JavaScript=enabled DTI data]).
:* Tri worked hard on making sure algorithm works on new high resolution 3T data (available here: [https://portal.nbirn.net/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=moveUpDir&gv7_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FPNL%2F3T_strct_dti_fmri%2Fcase01017&up=gv7&JavaScript=enabled|PNL 3T Data]).
:* Tests have been done also on the spiral diffusion phantom, to make sure diffusion directions and scanner coordinates are handled properly by the algorithm (Figure 6).

; B - Clinical Applications

:* Algorithm was used to trace and analyze anterior limb of the internal capsule on 1.5T data. It generated reacher representation of frontal fiber projections, it also turned out to be more sensitive to group differences in white matter integrity that conventional deterministic tractography (see Figure 6).
:* Algorithm was also used to trace smaller white matter fiber tracts, such as Cingulum, Fornix, Uncinate Fasciculus, Arcuate Fasciculus on 3T "Santa Fe" dataset (http://www.na-mic.org/Wiki/index.php/SanteFe.Tractography.Conference)

; C - References

:* [http://www.na-mic.org/Wiki/index.php/Image:IC_posternew.png Shenton, M.E., Ngo, T., Rosenberger, G., Westin, C.F., Levitt, J.J., McCarley, R.W., Kubicki, M. Study of Thalamo-Cortical White Matter Fiber Tract Projections in Schizophrenia Using Diffusion Stochastic Tractography. Poster presented at the 46th Meeting of the American College of Neuropsychopharmacology, Boca Raton, FL, December 2007.]
:* [http://www.na-mic.org/Wiki/index.php/Media:dougt_poster.pdf Terry DP, Rausch AC, Alvarado JL, Melonakos ED, Markant D, Westin CF, Kikinis R, von Siebenthal J, Shenton ME, Kubicki M. White Matter Properties of Emotion Related Connections in Schizophrenia. Poster presented at the 2009 Mysell Poster Day, Dept. of Psychiatry, Harvard Medical School, April 2009]
:* [http://www.na-mic.org/Wiki/index.php/Media:Jorge_poster.pdf Alvarado JL, Terry DP, Markant D, Ngo T, Kikinis R, Westin CF, McCarley RW, Shenton ME, Kubicki M. Study of Language-Related White Matter Tract Connections in Schizophrenia using Diffusion Stochastic Tractography. Poster presented at the 2009 Mysell Poster Day, Dept. of Psychiatry, Harvard Medical School, April 2009]
:* Melonakos ED, Shenton ME, Markant D, Alvarado J, Westin CF, Kubicki M. White Matter Properties of Orbitofrontal Connections in Schizophrenia. Poster being presented at the 64th Meeting of the Society of Biological Psychiatry. Vancouver, BC. May 2009.

== Work in Progress ==

; A - Optimization and Testing of stochastic tractography module :
:* Julien is testing now separate components of work flow, including the masks, and impact of their precision on stochastic output, as well as impact of number of seeding points on tractography results.
:* At the same time, we are testing the module on Max Plank dataset, as well as running it on tractography comparison project dataset.
:* We are discussing possibility of adding project specific functionality to module, such as third ROI to guide tractography, nonlinear registration button for merging fMRI and DTI data.

; B - Related Clinical Projects

;* Arcuate Fasciculus Extraction Project
[[Image:STArcuate.jpg|thumb|right|200px|<font size=1>Figure 7: The arcuate fasciculus including seed, midpoint and target ROI's.</font>]]
We have started the project of investigating Arcuate Fasciculus using Stochastic Tractography (Figure 7). This structure is especially important in both VCFS and schizophrenia, as it connects language related areas (Brocka and Wernicke's), and is involved in language processing quite disturbed in schizophrenia patients. It also can not be reliably traced using deterministic tractography.
:Project involves:
:* Whole brain segmentation, and automatic extraction of regions interconnected by Arcuate Fasciculus (Inferior frontal and Superior Temporal Gyri).
:* White matter segmentation, in order to prevent algorithm from traveling through the ventricles, where diffusivity is high.
:* Non-linear registration of labelmaps to the DTI space.
:* Seeding tracts. We have piloted it using 5000 seeds per voxel, however it is quite time consuming running it on even most powerful computers in the lab, so we have experimented with smaller number of seeds per voxel. We tested 1000 seeds, which gave virtually identical results.
:* Extracting path of interest, and calculating FA along the path for group comparison. Presentation of previous results for 7 schizophrenics and 12 control subjects, can be found here: [[Media:NAMIC_AHM_Arcuate.ppt|Progress Report Presentation]].
:* Results presentation and paper submission. Abstract was submitted and accepted for presentation at World Biological Psychiatry Symposium in Venice, Italy. Paper is in preparation.

;* Semantic Network Connectivity Project
We use combination of fMRI and DTI data to define and characterize functional and anatomical connectivity within the semantic processing network in schizophrenia.
:Project involves:
:* fMRI data analysis and identification of functional nodes involved in semantic processing in healthy controls and seubjects with schizophrenia
:* Analysis of functional connectivity (using FSL) between nodes of semantic network
:* Whole brain Voxel Based analysis of DTI data in same population
:* Use of stochastic tractography to identify connections between functional nodes
:* Correlational analysis involving anatomical and functional connectivity data.
:* Results presentation and paper submission. Data was presented at HBM conference in 2007, paper has been submitted to HBM.

;* Study of Default Network
We have started collaboration with Department of Neurophysiology Max Planck Institute in Frankfurt (Anna Rotarska contact person). They have dataset containing DTI and resting state fMRI in schizophrenia, and want to use stochastic tractography to measure integrity of anatomical connections within the default network in schizophrenia. Their fMRI data has been co-registered with anatomical scans, and we are putting it into the DTI space. Once this is done, we will start creating tracts connecting fMRI ROIs. This data will be also used to test robustness of our module, since data was collected on a different scanner (3T Philips).

;* Tractography Comparison Project
We are also working on a [http://www.na-mic.org/Wiki/index.php/SanteFe.Tractography.Conference tractography comparison project]dataset, where we apply stochastic tractography to phantom, as well as test dataset.

===Staffing Plan===

* Sylvain and Yogesh are the DBP resources charged with adapting the tools in the NA-MIC Kit to the DBP needs
* Doug Markant, our NAMIC RA has left the lab, and now Doug Terry, is a new NAMIC RA.
* Julien is our new NAMIC software engineer. He is responsible for improving the STM (Stochastic Tractography Module), and making sure software works with STM compliant datasets. [[Link Progress| Progress]]
* Polina is the algorithm core contact
* Brad is the engineering core contact

===Schedule===

* '''10/2007''' - Optimization of Stochastic Tractography algorythm for 1.5T data.
* '''10/2007''' - Algorythm testing on Santa Fe data set and diffusion phantom.
* '''06/2008''' - Optimization of Stochastic Tractography algorythm for 3T data.
* '''11/2008''' - Slicer 3 module prototype using python.
* '''12/2008''' - Slicer 3 module official release
* '''12/2008''' - Documentation and packaging for dissemination.
* '''12/2008''' - Arcuate Fasciculus results.
* '''01/2009''' - Arcuate Fasciuclus first draft of the paper.
* '''05/2009''' - Distortion correction and nonlinear registration added to the module
* '''05/2009''' - Symposium on tractography, including stochastic methods at World Biological Psychiatry Symposium in Florence, Italy.
* '''05/2009''' - Presentation of Arcuate Fasciculus findings at World Biological Psychiatry Symposium in Florence, Italy.

===Team and Institute===
*PI: Marek Kubicki (kubicki at bwh.harvard.edu)
*DBP2 Investigators: Sylvain Bouix, Yogesh Rathi, Julien de Siebenthal
*NA-MIC Engineering Contact: Brad Davis, Kitware
*NA-MIC Algorithms Contact: Polina Gollard, MIT

===Publications===

''In print''

*[http://www.na-mic.org/pages/Special:Publications?text=Kubicki+AND+Westin+AND+DTI&submit=Search&words=all&title=checked&keywords=checked&authors=checked&abstract=checked&sponsors=checked&searchbytag=checked| NA-MIC Publications Database - Clinical Applications]

*[http://www.na-mic.org/pages/Special:Publications?text=Ngo+AND+Golland&submit=Search&words=all&title=checked&keywords=checked&authors=checked&abstract=checked&sponsors=checked&searchbytag=checked| NA-MIC Publications Database - Algorithms Development]

[[Category: Schizophrenia]] [[Category: Diffusion MRI]] [[Category: Segmentation]]

DBP2:Harvard:Brain Segmentation Roadmap

2009-04-16T13:57:57Z

Dougt:

Back to [[NA-MIC_Internal_Collaborations|NA-MIC Collaborations]], [[DBP2:Harvard|Harvard DBP 2]]
__NOTOC__
=Stochastic Tractography for VCFS=
== Roadmap ==

The main goal of this project is to develop end-to-end application that would be used to characterize anatomical connectivity abnormalities in the brain of patients with velocardiofacial syndrome (VCFS), and to link this information with deficits in schizophrenia. This page describes the technology roadmap for stochastic tractography, using newly acquired 3T data, NAMIC tools and slicer 3.

== Algorithm ==

[[Image:IC_sto_new.png|thumb|right|200px|<font size=1> Figure 1: Comparison of deterministic and stochastic tractography algorithms</font>]]
; A-Description
* Most tractography methods estimate fibers by tracing the maximum direction of diffusion. A limitation of this approach is that, in practice, several factors introduce uncertainty in the tracking procedure, including, noise, splitting and crossing fibers, head motion and image artifacts. To address this uncertainty, stochastic tractography methods have been developed to quantify the uncertainty associated with estimated fibers (Bjornemo et al., 2002). Method uses a propagation model based on stochastics and regularization, which allows paths originating at one point to branch and return a probability distribution of possible paths. The method utilizes principles of a statistical Monte Carlo method called Sequential Importance Sampling and Resampling (SISR). Based on probability functions, using a sequential importance sampling technique ([http://lmi.bwh.harvard.edu/papers/pdfs/2002/bjornemoMICCAI02.pdf Bjornemo et al., 2002]), one can generate thousands of fibers starting in the same point by sequentially drawing random step directions. This gives a very rich model of the fiber distribution, as contrasted with single fibers produced by conventional tractography methods. Moreover, from a large number of sampled paths, probability maps can be generated, providing better estimates of connectivity between several anatomical locations. A comparison of the algorithms can be seen here. (Figure 1)

[[Image:StochasticPic.PNG|thumb|right|200px|<font size=1>Figure 2: Stochastic Tractography of uncinate fasciculis on anatomical data (left) and cingulum bungle on fMRI scan (right)</font>]]
; B-Possible Applications
* Since diffusion direction uncertainty within the gray matter is quite significant; principal diffusion direction approaches usually do not work for tracking between two gray matter regions. Thus if one requires finding connections between a priori selected anatomical gray matter regions, defined either by anatomical segmentations (in case of using structural ROI data), or functional activations (in case of megring DTI with fMRI), stochastic tractography seems to be the method of choice. Here is an example of this application to anatomical data (Figure 2, left image) and to fMRI data (Figure 2, right image).

* Stochastic Tractography is also comparable, if not better, in defining large white matter fiber bundles, especially those traveling through white matter regions characterized by increased diffusion uncertainty (fiber crossings). Example of such application to internal capsule. (Figure 3)

[[Image:IC-comp-new.png|thumb|right|200px|<font size=1>Figure 3: Streamline vs. Stochastic Tractography of the Internal Capsule</font>]]
; C-References

* [http://lmi.bwh.harvard.edu/papers/pdfs/2002/bjornemoMICCAI02.pdf Björnemo M, Brun A, Kikinis R, Westin CF. Regularized stochastic white matter tractography using diffusion tensor MRI. In Fifth International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI'02). Tokyo, Japan, 2002;435-442.]
* [http://lmi.bwh.harvard.edu/papers/pdfs/2006/frimanTMI06.pdf Friman, O., Farneback, G., Westin CF. A Bayesian Approach for Stochastic White Matter Tractography. IEEE Transactions on Medical Imaging, Vol 25, No. 8, Aug. 2006]

[[Image:stoch_menu.png|thumb|right|200px|<font size=1> Figure 4: Python Stochastic Tractography GUI </font>]]

==Module==
Can be found in: MODULES > PYTHON MODULES > PYTHON STOCHASTIC TRACTOGRAPHY
;Functionality of Python Stochastic Tractography module in Slicer 3.0
* IO:
Module reads files (DWI and ROIs) in nhdr format.
* Smoothing:
One can smooth the DWI data (only Gausian smoothing is supported at this time). We recommend it if the data is noisy.
* Brain Mask:
The Brain mask defines the volume in which the tensor will be computed and the tracts evaluated.
*Diffusion Tensor:
This step calculates the tensor and can output anisotropy indices (FA/Mode/Trace)
*Tractography:
Parameters that need to be adjusted:
:1. The amount of tracts that will be seeded from each voxel (we recommend between 500 and 1000 tracts, depending on the workstation power- 1000 tracts per voxel seeded within the large ROI for high resolution DWI can take a long time to compute).
:2. Maximum tract length (in mm), this can eliminate long, unwanted tracts if the regions for which connection is measured are located close to each other [[Image:stoch_menu2a.png|thumb|right|200px|<font size=1> Figure 5: Python Stochastic Tractography GUI, part 2 </font>]]
:3. Step Size: distance between each re-estimation of tensors, usually between 0.5 and 1 mm
:4. Stopping criteria. This can be used on the top of WM mask, to terminate tracts (in case they really want to travel through the CSF, for example).
*Connectivity Map:
This step creates output probability maps.
:1. binary: each voxel is counted only once if at least one fiber pass through it
:2. cumulative: tracts are summed by voxel independently
:3. weighted: tracts are summed by voxel depending on their length

Then, probability maps can be saved as ROIs, and either used directly, or thresholded (at certain probability, step claimed by few publications to remove noise) in slicer to mask and compute average FA, Mode, Trace for entire connection. Diffusion indices can be also weighted by the probability of connection for each voxel.

:* Module documentation can be found here:
:**[[Media:IJdata.tar.gz|Training Dataset]]
:**[[Media:PNLStochasictNew.ppt|Training Presentation]]
:**[[Media:Helix.zip|Sample Helix Dataset]]

==Work Accomplished==
[[Image:helix.png|thumb|right|200px|<font size=1>Figure 6: Stochastic Tractography on Phantom</font>]]
; A - Optimization and testing of stochastic tractography algorithm :
:* Original methodological paper, as well as our first attempts to use the algorithm (CC+ and matlab scripts) have been done on old "NAMIC" 1.5T LSDI data ([https://portal.nbirn.net:443/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=gotoDirectory&CPq_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FHarvard%2Farchive_morph&up=CPq&JavaScript=enabled Structural MRI] and [https://portal.nbirn.net:443/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=gotoDirectory&up=7li&7li_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FHarvard%2Farchive_diffusion&JavaScript=enabled DTI data]).
:* Tri worked hard on making sure algorithm works on new high resolution 3T data (available here: [https://portal.nbirn.net/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=moveUpDir&gv7_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FPNL%2F3T_strct_dti_fmri%2Fcase01017&up=gv7&JavaScript=enabled|PNL 3T Data]).
:* Tests have been done also on the spiral diffusion phantom, to make sure diffusion directions and scanner coordinates are handled properly by the algorithm (Figure 6).

; B - Clinical Applications

:* Algorithm was used to trace and analyze anterior limb of the internal capsule on 1.5T data. It generated reacher representation of frontal fiber projections, it also turned out to be more sensitive to group differences in white matter integrity that conventional deterministic tractography (see Figure 6).
:* Algorithm was also used to trace smaller white matter fiber tracts, such as Cingulum, Fornix, Uncinate Fasciculus, Arcuate Fasciculus on 3T "Santa Fe" dataset (http://www.na-mic.org/Wiki/index.php/SanteFe.Tractography.Conference)

; C - References

:* [http://www.na-mic.org/Wiki/index.php/Image:IC_posternew.png Shenton, M.E., Ngo, T., Rosenberger, G., Westin, C.F., Levitt, J.J., McCarley, R.W., Kubicki, M. Study of Thalamo-Cortical White Matter Fiber Tract Projections in Schizophrenia Using Diffusion Stochastic Tractography. Poster presented at the 46th Meeting of the American College of Neuropsychopharmacology, Boca Raton, FL, December 2007.]
:* [http://www.na-mic.org/Wiki/index.php/Media:dougt_poster.pdf Terry DP, Rausch AC, Alvarado JL, Melonakos ED, Markant D, Westin CF, Kikinis R, von Siebenthal J, Shenton ME, Kubicki M. White Matter Properties of Emotion Related Connections in Schizophrenia. Poster presented at the 2009 Mysell Poster Day, Dept. of Psychiatry, Harvard Medical School, April 2009]
:* [http://www.na-mic.org/Wiki/index.php/Media:Jorge_poster.pdf Alvarado JL, Terry DP, Markant D, Ngo T, Kikinis R, Westin CF, McCarley RW, Shenton ME, Kubicki M. Study of Language-Related White Matter Tract Connections in Schizophrenia using Diffusion Stochastic Tractography. Poster presented at the 2009 Mysell Poster Day, Dept. of Psychiatry, Harvard Medical School, April 2009]

== Work in Progress ==

; A - Optimization and Testing of stochastic tractography module :
:* Julien is testing now separate components of work flow, including the masks, and impact of their precision on stochastic output, as well as impact of number of seeding points on tractography results.
:* At the same time, we are testing the module on Max Plank dataset, as well as running it on tractography comparison project dataset.
:* We are discussing possibility of adding project specific functionality to module, such as third ROI to guide tractography, nonlinear registration button for merging fMRI and DTI data.

; B - Related Clinical Projects

;* Arcuate Fasciculus Extraction Project
[[Image:STArcuate.jpg|thumb|right|200px|<font size=1>Figure 7: The arcuate fasciculus including seed, midpoint and target ROI's.</font>]]
We have started the project of investigating Arcuate Fasciculus using Stochastic Tractography (Figure 7). This structure is especially important in both VCFS and schizophrenia, as it connects language related areas (Brocka and Wernicke's), and is involved in language processing quite disturbed in schizophrenia patients. It also can not be reliably traced using deterministic tractography.
:Project involves:
:* Whole brain segmentation, and automatic extraction of regions interconnected by Arcuate Fasciculus (Inferior frontal and Superior Temporal Gyri).
:* White matter segmentation, in order to prevent algorithm from traveling through the ventricles, where diffusivity is high.
:* Non-linear registration of labelmaps to the DTI space.
:* Seeding tracts. We have piloted it using 5000 seeds per voxel, however it is quite time consuming running it on even most powerful computers in the lab, so we have experimented with smaller number of seeds per voxel. We tested 1000 seeds, which gave virtually identical results.
:* Extracting path of interest, and calculating FA along the path for group comparison. Presentation of previous results for 7 schizophrenics and 12 control subjects, can be found here: [[Media:NAMIC_AHM_Arcuate.ppt|Progress Report Presentation]].
:* Results presentation and paper submission. Abstract was submitted and accepted for presentation at World Biological Psychiatry Symposium in Venice, Italy. Paper is in preparation.

;* Semantic Network Connectivity Project
We use combination of fMRI and DTI data to define and characterize functional and anatomical connectivity within the semantic processing network in schizophrenia.
:Project involves:
:* fMRI data analysis and identification of functional nodes involved in semantic processing in healthy controls and seubjects with schizophrenia
:* Analysis of functional connectivity (using FSL) between nodes of semantic network
:* Whole brain Voxel Based analysis of DTI data in same population
:* Use of stochastic tractography to identify connections between functional nodes
:* Correlational analysis involving anatomical and functional connectivity data.
:* Results presentation and paper submission. Data was presented at HBM conference in 2007, paper has been submitted to HBM.

;* Study of Default Network
We have started collaboration with Department of Neurophysiology Max Planck Institute in Frankfurt (Anna Rotarska contact person). They have dataset containing DTI and resting state fMRI in schizophrenia, and want to use stochastic tractography to measure integrity of anatomical connections within the default network in schizophrenia. Their fMRI data has been co-registered with anatomical scans, and we are putting it into the DTI space. Once this is done, we will start creating tracts connecting fMRI ROIs. This data will be also used to test robustness of our module, since data was collected on a different scanner (3T Philips).

;* Tractography Comparison Project
We are also working on a [http://www.na-mic.org/Wiki/index.php/SanteFe.Tractography.Conference tractography comparison project]dataset, where we apply stochastic tractography to phantom, as well as test dataset.

===Staffing Plan===

* Sylvain and Yogesh are the DBP resources charged with adapting the tools in the NA-MIC Kit to the DBP needs
* Doug Markant, our NAMIC RA has left the lab, and now Doug Terry, is a new NAMIC RA.
* Julien is our new NAMIC software engineer. He is responsible for improving the STM (Stochastic Tractography Module), and making sure software works with STM compliant datasets. [[Link Progress| Progress]]
* Polina is the algorithm core contact
* Brad is the engineering core contact

===Schedule===

* '''10/2007''' - Optimization of Stochastic Tractography algorythm for 1.5T data.
* '''10/2007''' - Algorythm testing on Santa Fe data set and diffusion phantom.
* '''06/2008''' - Optimization of Stochastic Tractography algorythm for 3T data.
* '''11/2008''' - Slicer 3 module prototype using python.
* '''12/2008''' - Slicer 3 module official release
* '''12/2008''' - Documentation and packaging for dissemination.
* '''12/2008''' - Arcuate Fasciculus results.
* '''01/2009''' - Arcuate Fasciuclus first draft of the paper.
* '''05/2009''' - Distortion correction and nonlinear registration added to the module
* '''05/2009''' - Symposium on tractography, including stochastic methods at World Biological Psychiatry Symposium in Florence, Italy.
* '''05/2009''' - Presentation of Arcuate Fasciculus findings at World Biological Psychiatry Symposium in Florence, Italy.

===Team and Institute===
*PI: Marek Kubicki (kubicki at bwh.harvard.edu)
*DBP2 Investigators: Sylvain Bouix, Yogesh Rathi, Julien de Siebenthal
*NA-MIC Engineering Contact: Brad Davis, Kitware
*NA-MIC Algorithms Contact: Polina Gollard, MIT

===Publications===

''In print''

*[http://www.na-mic.org/pages/Special:Publications?text=Kubicki+AND+Westin+AND+DTI&submit=Search&words=all&title=checked&keywords=checked&authors=checked&abstract=checked&sponsors=checked&searchbytag=checked| NA-MIC Publications Database - Clinical Applications]

*[http://www.na-mic.org/pages/Special:Publications?text=Ngo+AND+Golland&submit=Search&words=all&title=checked&keywords=checked&authors=checked&abstract=checked&sponsors=checked&searchbytag=checked| NA-MIC Publications Database - Algorithms Development]

[[Category: Schizophrenia]] [[Category: Diffusion MRI]] [[Category: Segmentation]]

DBP2:Harvard:Brain Segmentation Roadmap

2009-04-15T20:54:34Z

Dougt: /* Work Accomplished */

Back to [[NA-MIC_Internal_Collaborations|NA-MIC Collaborations]], [[DBP2:Harvard|Harvard DBP 2]]
__NOTOC__
=Stochastic Tractography for VCFS=
== Roadmap ==

The main goal of this project is to develop end-to-end application that would be used to characterize anatomical connectivity abnormalities in the brain of patients with velocardiofacial syndrome (VCFS), and to link this information with deficits in schizophrenia. This page describes the technology roadmap for stochastic tractography, using newly acquired 3T data, NAMIC tools and slicer 3.

== Algorithm ==

[[Image:IC_sto_new.png|thumb|right|200px|<font size=1> Figure 1: Comparison of deterministic and stochastic tractography algorithms</font>]]
; A-Description
* Most tractography methods estimate fibers by tracing the maximum direction of diffusion. A limitation of this approach is that, in practice, several factors introduce uncertainty in the tracking procedure, including, noise, splitting and crossing fibers, head motion and image artifacts. To address this uncertainty, stochastic tractography methods have been developed to quantify the uncertainty associated with estimated fibers (Bjornemo et al., 2002). Method uses a propagation model based on stochastics and regularization, which allows paths originating at one point to branch and return a probability distribution of possible paths. The method utilizes principles of a statistical Monte Carlo method called Sequential Importance Sampling and Resampling (SISR). Based on probability functions, using a sequential importance sampling technique ([http://lmi.bwh.harvard.edu/papers/pdfs/2002/bjornemoMICCAI02.pdf Bjornemo et al., 2002]), one can generate thousands of fibers starting in the same point by sequentially drawing random step directions. This gives a very rich model of the fiber distribution, as contrasted with single fibers produced by conventional tractography methods. Moreover, from a large number of sampled paths, probability maps can be generated, providing better estimates of connectivity between several anatomical locations. A comparison of the algorithms can be seen here. (Figure 1)

[[Image:StochasticPic.PNG|thumb|right|200px|<font size=1>Figure 2: Stochastic Tractography of uncinate fasciculis on anatomical data (left) and cingulum bungle on fMRI scan (right)</font>]]
; B-Possible Applications
* Since diffusion direction uncertainty within the gray matter is quite significant; principal diffusion direction approaches usually do not work for tracking between two gray matter regions. Thus if one requires finding connections between a priori selected anatomical gray matter regions, defined either by anatomical segmentations (in case of using structural ROI data), or functional activations (in case of megring DTI with fMRI), stochastic tractography seems to be the method of choice. Here is an example of this application to anatomical data (Figure 2, left image) and to fMRI data (Figure 2, right image).

* Stochastic Tractography is also comparable, if not better, in defining large white matter fiber bundles, especially those traveling through white matter regions characterized by increased diffusion uncertainty (fiber crossings). Example of such application to internal capsule. (Figure 3)

[[Image:IC-comp-new.png|thumb|right|200px|<font size=1>Figure 3: Streamline vs. Stochastic Tractography of the Internal Capsule</font>]]
; C-References

* [http://lmi.bwh.harvard.edu/papers/pdfs/2002/bjornemoMICCAI02.pdf Björnemo M, Brun A, Kikinis R, Westin CF. Regularized stochastic white matter tractography using diffusion tensor MRI. In Fifth International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI'02). Tokyo, Japan, 2002;435-442.]
* [http://lmi.bwh.harvard.edu/papers/pdfs/2006/frimanTMI06.pdf Friman, O., Farneback, G., Westin CF. A Bayesian Approach for Stochastic White Matter Tractography. IEEE Transactions on Medical Imaging, Vol 25, No. 8, Aug. 2006]

[[Image:stoch_menu.png|thumb|right|200px|<font size=1> Figure 4: Python Stochastic Tractography GUI </font>]]

==Module==
Can be found in: MODULES > PYTHON MODULES > PYTHON STOCHASTIC TRACTOGRAPHY
;Functionality of Python Stochastic Tractography module in Slicer 3.0
* IO:
Module reads files (DWI and ROIs) in nhdr format.
* Smoothing:
One can smooth the DWI data (only Gausian smoothing is supported at this time). We recommend it if the data is noisy.
* Brain Mask:
The Brain mask defines the volume in which the tensor will be computed and the tracts evaluated.
*Diffusion Tensor:
This step calculates the tensor and can output anisotropy indices (FA/Mode/Trace)
*Tractography:
Parameters that need to be adjusted:
:1. The amount of tracts that will be seeded from each voxel (we recommend between 500 and 1000 tracts, depending on the workstation power- 1000 tracts per voxel seeded within the large ROI for high resolution DWI can take a long time to compute).
:2. Maximum tract length (in mm), this can eliminate long, unwanted tracts if the regions for which connection is measured are located close to each other [[Image:stoch_menu2a.png|thumb|right|200px|<font size=1> Figure 5: Python Stochastic Tractography GUI, part 2 </font>]]
:3. Step Size: distance between each re-estimation of tensors, usually between 0.5 and 1 mm
:4. Stopping criteria. This can be used on the top of WM mask, to terminate tracts (in case they really want to travel through the CSF, for example).
*Connectivity Map:
This step creates output probability maps.
:1. binary: each voxel is counted only once if at least one fiber pass through it
:2. cumulative: tracts are summed by voxel independently
:3. weighted: tracts are summed by voxel depending on their length

Then, probability maps can be saved as ROIs, and either used directly, or thresholded (at certain probability, step claimed by few publications to remove noise) in slicer to mask and compute average FA, Mode, Trace for entire connection. Diffusion indices can be also weighted by the probability of connection for each voxel.

:* Module documentation can be found here:
:**[[Media:IJdata.tar.gz|Training Dataset]]
:**[[Media:PNLStochasictNew.ppt|Training Presentation]]
:**[[Media:Helix.zip|Sample Helix Dataset]]

==Work Accomplished==
[[Image:helix.png|thumb|right|200px|<font size=1>Figure 6: Stochastic Tractography on Phantom</font>]]
; A - Optimization and testing of stochastic tractography algorithm :
:* Original methodological paper, as well as our first attempts to use the algorithm (CC+ and matlab scripts) have been done on old "NAMIC" 1.5T LSDI data ([https://portal.nbirn.net:443/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=gotoDirectory&CPq_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FHarvard%2Farchive_morph&up=CPq&JavaScript=enabled Structural MRI] and [https://portal.nbirn.net:443/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=gotoDirectory&up=7li&7li_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FHarvard%2Farchive_diffusion&JavaScript=enabled DTI data]).
:* Tri worked hard on making sure algorithm works on new high resolution 3T data (available here: [https://portal.nbirn.net/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=moveUpDir&gv7_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FPNL%2F3T_strct_dti_fmri%2Fcase01017&up=gv7&JavaScript=enabled|PNL 3T Data]).
:* Tests have been done also on the spiral diffusion phantom, to make sure diffusion directions and scanner coordinates are handled properly by the algorithm (Figure 6).

; B - Clincal Applications

:* Algorithm was used to trace and analyze anterior limb of the internal capsule on 1.5T data. It generated reacher representation of frontal fiber projections, it also turned out to be more sensitive to group differences in white matter integrity that conventional deterministic tractography (see Figure 6).
:* Algorithm was also used to trace smaller white matter fiber tracts, such as Cingulum, Fornix, Uncinate Fasciculus, Arcuate Fasciculus on 3T "Santa Fe" dataset (http://www.na-mic.org/Wiki/index.php/SanteFe.Tractography.Conference)

; C - References

:* [http://www.na-mic.org/Wiki/index.php/Image:IC_posternew.png Shenton, M.E., Ngo, T., Rosenberger, G., Westin, C.F., Levitt, J.J., McCarley, R.W., Kubicki, M. Study of Thalamo-Cortical White Matter Fiber Tract Projections in Schizophrenia Using Diffusion Stochastic Tractography. Poster presented at the 46th Meeting of the American College of Neuropsychopharmacology, Boca Raton, FL, December 2007.]
:* [http://www.na-mic.org/Wiki/index.php/Media:dougt_poster.pdf Terry DP, Rausch AC, Alvarado JL, Melonakos ED, Markant D, Westin CF, Kikinis R, von Siebenthal J, Shenton ME, Kubicki M. White Matter Properties of Emotion Related Connections in Schizophrenia. Poster presented at the 2009 Mysell Poster Day, Dept. of Psychiatry, Harvard Medical School, April 2009]
:* [http://www.na-mic.org/Wiki/index.php/Media:Jorge_poster.pdf Alvarado JL, Terry DP, Markant D, Ngo T, Kikinis R, Westin CF, McCarley RW, Shenton ME, Kubicki M. Study of Language-Related White Matter Tract Connections in Schizophrenia using Diffusion Stochastic Tractography. Poster presented at the 2009 Mysell Poster Day, Dept. of Psychiatry, Harvard Medical School, April 2009]

== Work in Progress ==

; A - Optimization and Testing of stochastic tractography module :
:* Julien is testing now separate components of work flow, including the masks, and impact of their precision on stochastic output, as well as impact of number of seeding points on tractography results.
:* At the same time, we are testing the module on Max Plank dataset, as well as running it on tractography comparison project dataset.
:* We are discussing possibility of adding project specific functionality to module, such as third ROI to guide tractography, nonlinear registration button for merging fMRI and DTI data.

; B - Related Clinical Projects

;* Arcuate Fasciculus Extraction Project
[[Image:STArcuate.jpg|thumb|right|200px|<font size=1>Figure 7: The arcuate fasciculus including seed, midpoint and target ROI's.</font>]]
We have started the project of investigating Arcuate Fasciculus using Stochastic Tractography (Figure 7). This structure is especially important in both VCFS and schizophrenia, as it connects language related areas (Brocka and Wernicke's), and is involved in language processing quite disturbed in schizophrenia patients. It also can not be reliably traced using deterministic tractography.
:Project involves:
:* Whole brain segmentation, and automatic extraction of regions interconnected by Arcuate Fasciculus (Inferior frontal and Superior Temporal Gyri).
:* White matter segmentation, in order to prevent algorithm from traveling through the ventricles, where diffusivity is high.
:* Non-linear registration of labelmaps to the DTI space.
:* Seeding tracts. We have piloted it using 5000 seeds per voxel, however it is quite time consuming running it on even most powerful computers in the lab, so we have experimented with smaller number of seeds per voxel. We tested 1000 seeds, which gave virtually identical results.
:* Extracting path of interest, and calculating FA along the path for group comparison. Presentation of previous results for 7 schizophrenics and 12 control subjects, can be found here: [[Media:NAMIC_AHM_Arcuate.ppt|Progress Report Presentation]].
:* Results presentation and paper submission. Abstract was submitted and accepted for presentation at World Biological Psychiatry Symposium in Venice, Italy. Paper is in preparation.

;* Semantic Network Connectivity Project
We use combination of fMRI and DTI data to define and characterize functional and anatomical connectivity within the semantic processing network in schizophrenia.
:Project involves:
:* fMRI data analysis and identification of functional nodes involved in semantic processing in healthy controls and seubjects with schizophrenia
:* Analysis of functional connectivity (using FSL) between nodes of semantic network
:* Whole brain Voxel Based analysis of DTI data in same population
:* Use of stochastic tractography to identify connections between functional nodes
:* Correlational analysis involving anatomical and functional connectivity data.
:* Results presentation and paper submission. Data was presented at HBM conference in 2007, paper has been submitted to HBM.

;* Study of Default Network
We have started collaboration with Department of Neurophysiology Max Planck Institute in Frankfurt (Anna Rotarska contact person). They have dataset containing DTI and resting state fMRI in schizophrenia, and want to use stochastic tractography to measure integrity of anatomical connections within the default network in schizophrenia. Their fMRI data has been co-registered with anatomical scans, and we are putting it into the DTI space. Once this is done, we will start creating tracts connecting fMRI ROIs. This data will be also used to test robustness of our module, since data was collected on a different scanner (3T Philips).

;* Tractography Comparison Project
We are also working on a [http://www.na-mic.org/Wiki/index.php/SanteFe.Tractography.Conference tractography comparison project]dataset, where we apply stochastic tractography to phantom, as well as test dataset.

===Staffing Plan===

* Sylvain and Yogesh are the DBP resources charged with adapting the tools in the NA-MIC Kit to the DBP needs
* Doug Markant, our NAMIC RA has left the lab, and now Doug Terry, is a new NAMIC RA.
* Julien is our new NAMIC software engineer. He is responsible for improving the STM (Stochastic Tractography Module), and making sure software works with STM compliant datasets. [[Link Progress| Progress]]
* Polina is the algorithm core contact
* Brad is the engineering core contact

===Schedule===

* '''10/2007''' - Optimization of Stochastic Tractography algorythm for 1.5T data.
* '''10/2007''' - Algorythm testing on Santa Fe data set and diffusion phantom.
* '''06/2008''' - Optimization of Stochastic Tractography algorythm for 3T data.
* '''11/2008''' - Slicer 3 module prototype using python.
* '''12/2008''' - Slicer 3 module official release
* '''12/2008''' - Documentation and packaging for dissemination.
* '''12/2008''' - Arcuate Fasciculus results.
* '''01/2009''' - Arcuate Fasciuclus first draft of the paper.
* '''05/2009''' - Distortion correction and nonlinear registration added to the module
* '''05/2009''' - Symposium on tractography, including stochastic methods at World Biological Psychiatry Symposium in Florence, Italy.
* '''05/2009''' - Presentation of Arcuate Fasciculus findings at World Biological Psychiatry Symposium in Florence, Italy.

===Team and Institute===
*PI: Marek Kubicki (kubicki at bwh.harvard.edu)
*DBP2 Investigators: Sylvain Bouix, Yogesh Rathi, Julien de Siebenthal
*NA-MIC Engineering Contact: Brad Davis, Kitware
*NA-MIC Algorithms Contact: Polina Gollard, MIT

===Publications===

''In print''

*[http://www.na-mic.org/pages/Special:Publications?text=Kubicki+AND+Westin+AND+DTI&submit=Search&words=all&title=checked&keywords=checked&authors=checked&abstract=checked&sponsors=checked&searchbytag=checked| NA-MIC Publications Database - Clinical Applications]

*[http://www.na-mic.org/pages/Special:Publications?text=Ngo+AND+Golland&submit=Search&words=all&title=checked&keywords=checked&authors=checked&abstract=checked&sponsors=checked&searchbytag=checked| NA-MIC Publications Database - Algorithms Development]

[[Category: Schizophrenia]] [[Category: Diffusion MRI]] [[Category: Segmentation]]

File:Jorge poster.pdf

2009-04-15T20:29:26Z

Dougt:

File:Jorge poster.ppt

2009-04-15T18:30:05Z

Dougt:

File:Dougt poster.pdf

2009-04-15T18:18:44Z

Dougt:

DBP2:Harvard:Brain Segmentation Roadmap

2009-04-15T18:11:47Z

Dougt:

Back to [[NA-MIC_Internal_Collaborations|NA-MIC Collaborations]], [[DBP2:Harvard|Harvard DBP 2]]
__NOTOC__
=Stochastic Tractography for VCFS=
== Roadmap ==

The main goal of this project is to develop end-to-end application that would be used to characterize anatomical connectivity abnormalities in the brain of patients with velocardiofacial syndrome (VCFS), and to link this information with deficits in schizophrenia. This page describes the technology roadmap for stochastic tractography, using newly acquired 3T data, NAMIC tools and slicer 3.

== Algorithm ==

[[Image:IC_sto_new.png|thumb|right|200px|<font size=1> Figure 1: Comparison of deterministic and stochastic tractography algorithms</font>]]
; A-Description
* Most tractography methods estimate fibers by tracing the maximum direction of diffusion. A limitation of this approach is that, in practice, several factors introduce uncertainty in the tracking procedure, including, noise, splitting and crossing fibers, head motion and image artifacts. To address this uncertainty, stochastic tractography methods have been developed to quantify the uncertainty associated with estimated fibers (Bjornemo et al., 2002). Method uses a propagation model based on stochastics and regularization, which allows paths originating at one point to branch and return a probability distribution of possible paths. The method utilizes principles of a statistical Monte Carlo method called Sequential Importance Sampling and Resampling (SISR). Based on probability functions, using a sequential importance sampling technique ([http://lmi.bwh.harvard.edu/papers/pdfs/2002/bjornemoMICCAI02.pdf Bjornemo et al., 2002]), one can generate thousands of fibers starting in the same point by sequentially drawing random step directions. This gives a very rich model of the fiber distribution, as contrasted with single fibers produced by conventional tractography methods. Moreover, from a large number of sampled paths, probability maps can be generated, providing better estimates of connectivity between several anatomical locations. A comparison of the algorithms can be seen here. (Figure 1)

[[Image:StochasticPic.PNG|thumb|right|200px|<font size=1>Figure 2: Stochastic Tractography of uncinate fasciculis on anatomical data (left) and cingulum bungle on fMRI scan (right)</font>]]
; B-Possible Applications
* Since diffusion direction uncertainty within the gray matter is quite significant; principal diffusion direction approaches usually do not work for tracking between two gray matter regions. Thus if one requires finding connections between a priori selected anatomical gray matter regions, defined either by anatomical segmentations (in case of using structural ROI data), or functional activations (in case of megring DTI with fMRI), stochastic tractography seems to be the method of choice. Here is an example of this application to anatomical data (Figure 2, left image) and to fMRI data (Figure 2, right image).

* Stochastic Tractography is also comparable, if not better, in defining large white matter fiber bundles, especially those traveling through white matter regions characterized by increased diffusion uncertainty (fiber crossings). Example of such application to internal capsule. (Figure 3)

[[Image:IC-comp-new.png|thumb|right|200px|<font size=1>Figure 3: Streamline vs. Stochastic Tractography of the Internal Capsule</font>]]
; C-References

* [http://lmi.bwh.harvard.edu/papers/pdfs/2002/bjornemoMICCAI02.pdf Björnemo M, Brun A, Kikinis R, Westin CF. Regularized stochastic white matter tractography using diffusion tensor MRI. In Fifth International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI'02). Tokyo, Japan, 2002;435-442.]
* [http://lmi.bwh.harvard.edu/papers/pdfs/2006/frimanTMI06.pdf Friman, O., Farneback, G., Westin CF. A Bayesian Approach for Stochastic White Matter Tractography. IEEE Transactions on Medical Imaging, Vol 25, No. 8, Aug. 2006]

[[Image:stoch_menu.png|thumb|right|200px|<font size=1> Figure 4: Python Stochastic Tractography GUI </font>]]

==Module==
Can be found in: MODULES > PYTHON MODULES > PYTHON STOCHASTIC TRACTOGRAPHY
;Functionality of Python Stochastic Tractography module in Slicer 3.0
* IO:
Module reads files (DWI and ROIs) in nhdr format.
* Smoothing:
One can smooth the DWI data (only Gausian smoothing is supported at this time). We recommend it if the data is noisy.
* Brain Mask:
The Brain mask defines the volume in which the tensor will be computed and the tracts evaluated.
*Diffusion Tensor:
This step calculates the tensor and can output anisotropy indices (FA/Mode/Trace)
*Tractography:
Parameters that need to be adjusted:
:1. The amount of tracts that will be seeded from each voxel (we recommend between 500 and 1000 tracts, depending on the workstation power- 1000 tracts per voxel seeded within the large ROI for high resolution DWI can take a long time to compute).
:2. Maximum tract length (in mm), this can eliminate long, unwanted tracts if the regions for which connection is measured are located close to each other [[Image:stoch_menu2a.png|thumb|right|200px|<font size=1> Figure 5: Python Stochastic Tractography GUI, part 2 </font>]]
:3. Step Size: distance between each re-estimation of tensors, usually between 0.5 and 1 mm
:4. Stopping criteria. This can be used on the top of WM mask, to terminate tracts (in case they really want to travel through the CSF, for example).
*Connectivity Map:
This step creates output probability maps.
:1. binary: each voxel is counted only once if at least one fiber pass through it
:2. cumulative: tracts are summed by voxel independently
:3. weighted: tracts are summed by voxel depending on their length

Then, probability maps can be saved as ROIs, and either used directly, or thresholded (at certain probability, step claimed by few publications to remove noise) in slicer to mask and compute average FA, Mode, Trace for entire connection. Diffusion indices can be also weighted by the probability of connection for each voxel.

:* Module documentation can be found here:
:**[[Media:IJdata.tar.gz|Training Dataset]]
:**[[Media:PNLStochasictNew.ppt|Training Presentation]]
:**[[Media:Helix.zip|Sample Helix Dataset]]

==Work Accomplished==
[[Image:helix.png|thumb|right|200px|<font size=1>Figure 6: Stochastic Tractography on Phantom</font>]]
; A - Optimization and testing of stochastic tractography algorithm :
:* Original methodological paper, as well as our first attempts to use the algorithm (CC+ and matlab scripts) have been done on old "NAMIC" 1.5T LSDI data ([https://portal.nbirn.net:443/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=gotoDirectory&CPq_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FHarvard%2Farchive_morph&up=CPq&JavaScript=enabled Structural MRI] and [https://portal.nbirn.net:443/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=gotoDirectory&up=7li&7li_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FHarvard%2Farchive_diffusion&JavaScript=enabled DTI data]).
:* Tri worked hard on making sure algorithm works on new high resolution 3T data (available here: [https://portal.nbirn.net/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=moveUpDir&gv7_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FPNL%2F3T_strct_dti_fmri%2Fcase01017&up=gv7&JavaScript=enabled|PNL 3T Data]).
:* Tests have been done also on the spiral diffusion phantom, to make sure diffusion directions and scanner coordinates are handled properly by the algorithm (Figure 6).

; B - Clincal Applications

:* Algorithm was used to trace and analyze anterior limb of the internal capsule on 1.5T data. It generated reacher representation of frontal fiber projections, it also turned out to be more sensitive to group differences in white matter integrity that conventional deterministic tractography (see Figure 6).
:* Algorithm was also used to trace smaller white matter fiber tracts, such as Cingulum, Fornix, Uncinate Fasciculus, Arcuate Fasciculus on 3T "Santa Fe" dataset (http://www.na-mic.org/Wiki/index.php/SanteFe.Tractography.Conference)

; C - References

:* [http://www.na-mic.org/Wiki/index.php/Image:IC_posternew.png Shenton, M.E., Ngo, T., Rosenberger, G., Westin, C.F., Levitt, J.J., McCarley, R.W., Kubicki, M. Study of Thalamo-Cortical White Matter Fiber Tract Projections in Schizophrenia Using Diffusion Stochastic Tractography. Poster presented at the 46th Meeting of the American College of Neuropsychopharmacology, Boca Raton, FL, December 2007.]

== Work in Progress ==

; A - Optimization and Testing of stochastic tractography module :
:* Julien is testing now separate components of work flow, including the masks, and impact of their precision on stochastic output, as well as impact of number of seeding points on tractography results.
:* At the same time, we are testing the module on Max Plank dataset, as well as running it on tractography comparison project dataset.
:* We are discussing possibility of adding project specific functionality to module, such as third ROI to guide tractography, nonlinear registration button for merging fMRI and DTI data.

; B - Related Clinical Projects

;* Arcuate Fasciculus Extraction Project
[[Image:STArcuate.jpg|thumb|right|200px|<font size=1>Figure 7: The arcuate fasciculus including seed, midpoint and target ROI's.</font>]]
We have started the project of investigating Arcuate Fasciculus using Stochastic Tractography (Figure 7). This structure is especially important in both VCFS and schizophrenia, as it connects language related areas (Brocka and Wernicke's), and is involved in language processing quite disturbed in schizophrenia patients. It also can not be reliably traced using deterministic tractography.
:Project involves:
:* Whole brain segmentation, and automatic extraction of regions interconnected by Arcuate Fasciculus (Inferior frontal and Superior Temporal Gyri).
:* White matter segmentation, in order to prevent algorithm from traveling through the ventricles, where diffusivity is high.
:* Non-linear registration of labelmaps to the DTI space.
:* Seeding tracts. We have piloted it using 5000 seeds per voxel, however it is quite time consuming running it on even most powerful computers in the lab, so we have experimented with smaller number of seeds per voxel. We tested 1000 seeds, which gave virtually identical results.
:* Extracting path of interest, and calculating FA along the path for group comparison. Presentation of previous results for 7 schizophrenics and 12 control subjects, can be found here: [[Media:NAMIC_AHM_Arcuate.ppt|Progress Report Presentation]].
:* Results presentation and paper submission. Abstract was submitted and accepted for presentation at World Biological Psychiatry Symposium in Venice, Italy. Paper is in preparation.

;* Semantic Network Connectivity Project
We use combination of fMRI and DTI data to define and characterize functional and anatomical connectivity within the semantic processing network in schizophrenia.
:Project involves:
:* fMRI data analysis and identification of functional nodes involved in semantic processing in healthy controls and seubjects with schizophrenia
:* Analysis of functional connectivity (using FSL) between nodes of semantic network
:* Whole brain Voxel Based analysis of DTI data in same population
:* Use of stochastic tractography to identify connections between functional nodes
:* Correlational analysis involving anatomical and functional connectivity data.
:* Results presentation and paper submission. Data was presented at HBM conference in 2007, paper has been submitted to HBM.

;* Study of Default Network
We have started collaboration with Department of Neurophysiology Max Planck Institute in Frankfurt (Anna Rotarska contact person). They have dataset containing DTI and resting state fMRI in schizophrenia, and want to use stochastic tractography to measure integrity of anatomical connections within the default network in schizophrenia. Their fMRI data has been co-registered with anatomical scans, and we are putting it into the DTI space. Once this is done, we will start creating tracts connecting fMRI ROIs. This data will be also used to test robustness of our module, since data was collected on a different scanner (3T Philips).

;* Tractography Comparison Project
We are also working on a [http://www.na-mic.org/Wiki/index.php/SanteFe.Tractography.Conference tractography comparison project]dataset, where we apply stochastic tractography to phantom, as well as test dataset.

===Staffing Plan===

* Sylvain and Yogesh are the DBP resources charged with adapting the tools in the NA-MIC Kit to the DBP needs
* Doug Markant, our NAMIC RA has left the lab, and now Doug Terry, is a new NAMIC RA.
* Julien is our new NAMIC software engineer. He is responsible for improving the STM (Stochastic Tractography Module), and making sure software works with STM compliant datasets. [[Link Progress| Progress]]
* Polina is the algorithm core contact
* Brad is the engineering core contact

===Schedule===

* '''10/2007''' - Optimization of Stochastic Tractography algorythm for 1.5T data.
* '''10/2007''' - Algorythm testing on Santa Fe data set and diffusion phantom.
* '''06/2008''' - Optimization of Stochastic Tractography algorythm for 3T data.
* '''11/2008''' - Slicer 3 module prototype using python.
* '''12/2008''' - Slicer 3 module official release
* '''12/2008''' - Documentation and packaging for dissemination.
* '''12/2008''' - Arcuate Fasciculus results.
* '''01/2009''' - Arcuate Fasciuclus first draft of the paper.
* '''05/2009''' - Distortion correction and nonlinear registration added to the module
* '''05/2009''' - Symposium on tractography, including stochastic methods at World Biological Psychiatry Symposium in Florence, Italy.
* '''05/2009''' - Presentation of Arcuate Fasciculus findings at World Biological Psychiatry Symposium in Florence, Italy.

===Team and Institute===
*PI: Marek Kubicki (kubicki at bwh.harvard.edu)
*DBP2 Investigators: Sylvain Bouix, Yogesh Rathi, Julien de Siebenthal
*NA-MIC Engineering Contact: Brad Davis, Kitware
*NA-MIC Algorithms Contact: Polina Gollard, MIT

===Publications===

''In print''

*[http://www.na-mic.org/pages/Special:Publications?text=Kubicki+AND+Westin+AND+DTI&submit=Search&words=all&title=checked&keywords=checked&authors=checked&abstract=checked&sponsors=checked&searchbytag=checked| NA-MIC Publications Database - Clinical Applications]

*[http://www.na-mic.org/pages/Special:Publications?text=Ngo+AND+Golland&submit=Search&words=all&title=checked&keywords=checked&authors=checked&abstract=checked&sponsors=checked&searchbytag=checked| NA-MIC Publications Database - Algorithms Development]

[[Category: Schizophrenia]] [[Category: Diffusion MRI]] [[Category: Segmentation]]

DBP2:Harvard:Brain Segmentation Roadmap

2009-04-15T16:24:47Z

Dougt: /* Stochastic Tractography for VCFS */

Back to [[NA-MIC_Internal_Collaborations|NA-MIC Collaborations]], [[DBP2:Harvard|Harvard DBP 2]]
__NOTOC__
=Stochastic Tractography for VCFS=
== Roadmap ==

The main goal of this project is to develop end-to-end application that would be used to characterize anatomical connectivity abnormalities in the brain of patients with velocardiofacial syndrome (VCFS), and to link this information with deficits in schizophrenia. This page describes the technology roadmap for stochastic tractography, using newly acquired 3T data, NAMIC tools and slicer 3.

== Algorithm ==

{|cellpadding="0"

| style="width:85%" |

; A-Description
* Most tractography methods estimate fibers by tracing the maximum direction of diffusion. A limitation of this approach is that, in practice, several factors introduce uncertainty in the tracking procedure, including, noise, splitting and crossing fibers, head motion and image artifacts. To address this uncertainty, stochastic tractography methods have been developed to quantify the uncertainty associated with estimated fibers (Bjornemo et al., 2002). Method uses a propagation model based on stochastics and regularization, which allows paths originating at one point to branch and return a probability distribution of possible paths. The method utilizes principles of a statistical Monte Carlo method called Sequential Importance Sampling and Resampling (SISR). Based on probability functions, using a sequential importance sampling technique ([http://lmi.bwh.harvard.edu/papers/pdfs/2002/bjornemoMICCAI02.pdf Bjornemo et al., 2002]), one can generate thousands of fibers starting in the same point by sequentially drawing random step directions. This gives a very rich model of the fiber distribution, as contrasted with single fibers produced by conventional tractography methods. Moreover, from a large number of sampled paths, probability maps can be generated, providing better estimates of connectivity between several anatomical locations. A comparison of the algorithms can be seen here. (Figure 1)

| style="width:15%" | [[Image:IC_sto_new.png|thumb|right|200px|<font size=1> Figure 1: Comparison of deterministic and stochastic tractography algorithms</font>]]

|}
{|cellpadding="0"

| style="width:85%" |

; B-Possible Applications
* Since diffusion direction uncertainty within the gray matter is quite significant; principal diffusion direction approaches usually do not work for tracking between two gray matter regions. Thus if one requires finding connections between a priori selected anatomical gray matter regions, defined either by anatomical segmentations (in case of using structural ROI data), or functional activations (in case of megring DTI with fMRI), stochastic tractography seems to be the method of choice. Here is an example of this application to anatomical data (Figure 2, left image) and to fMRI data (Figure 2, right image).

* Stochastic Tractography is also comparable, if not better, in defining large white matter fiber bundles, especially those traveling through white matter regions characterized by increased diffusion uncertainty (fiber crossings). Example of such application to internal capsule. (Figure 3)

| style="width:15%" |[[Image:StochasticPic.PNG|thumb|right|200px|<font size=1>Figure 2: Stochastic Tractography of uncinate fasciculis on anatomical data (left) and cingulum bungle on fMRI scan (right)</font>]]
|} [[Image:IC-comp-new.png|thumb|right|200px|<font size=1>Figure 3: Streamline vs. Stochastic Tractography of the Internal Capsule</font>]]
; C-References

* [http://lmi.bwh.harvard.edu/papers/pdfs/2002/bjornemoMICCAI02.pdf Björnemo M, Brun A, Kikinis R, Westin CF. Regularized stochastic white matter tractography using diffusion tensor MRI. In Fifth International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI'02). Tokyo, Japan, 2002;435-442.]
* [http://lmi.bwh.harvard.edu/papers/pdfs/2006/frimanTMI06.pdf Friman, O., Farneback, G., Westin CF. A Bayesian Approach for Stochastic White Matter Tractography. IEEE Transactions on Medical Imaging, Vol 25, No. 8, Aug. 2006]

[[Image:stoch_menu.png|thumb|right|200px|<font size=1> Figure 4: Python Stochastic Tractography GUI </font>]]

==Module==
Can be found in: MODULES > PYTHON MODULES > PYTHON STOCHASTIC TRACTOGRAPHY
;Functionality of Python Stochastic Tractography module in Slicer 3.0
* IO:
Module reads files (DWI and ROIs) in nhdr format.
* Smoothing:
One can smooth the DWI data (only Gausian smoothing is supported at this time). We recommend it if the data is noisy.
* Brain Mask:
The Brain mask defines the volume in which the tensor will be computed and the tracts evaluated.
*Diffusion Tensor:
This step calculates the tensor and can output anisotropy indices (FA/Mode/Trace)
*Tractography:
Parameters that need to be adjusted:
:1. The amount of tracts that will be seeded from each voxel (we recommend between 500 and 1000 tracts, depending on the workstation power- 1000 tracts per voxel seeded within the large ROI for high resolution DWI can take a long time to compute).
:2. Maximum tract length (in mm), this can eliminate long, unwanted tracts if the regions for which connection is measured are located close to each other [[Image:stoch_menu2a.png|thumb|right|200px|<font size=1> Figure 5: Python Stochastic Tractography GUI, part 2 </font>]]
:3. Step Size: distance between each re-estimation of tensors, usually between 0.5 and 1 mm
:4. Stopping criteria. This can be used on the top of WM mask, to terminate tracts (in case they really want to travel through the CSF, for example).
*Connectivity Map:
This step creates output probability maps.
:1. binary: each voxel is counted only once if at least one fiber pass through it
:2. cumulative: tracts are summed by voxel independently
:3. weighted: tracts are summed by voxel depending on their length

Then, probability maps can be saved as ROIs, and either used directly, or thresholded (at certain probability, step claimed by few publications to remove noise) in slicer to mask and compute average FA, Mode, Trace for entire connection. Diffusion indices can be also weighted by the probability of connection for each voxel.

:* Module documentation can be found here:
:**[[Media:IJdata.tar.gz|Training Dataset]]
:**[[Media:PNLStochasictNew.ppt|Training Presentation]]
:**[[Media:Helix.zip|Sample Helix Dataset]]

==Work Accomplished==
[[Image:helix.png|thumb|right|200px|<font size=1>Figure 6: Stochastic Tractography on Phantom</font>]]
; A - Optimization and testing of stochastic tractography algorithm :
:* Original methodological paper, as well as our first attempts to use the algorithm (CC+ and matlab scripts) have been done on old "NAMIC" 1.5T LSDI data ([https://portal.nbirn.net:443/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=gotoDirectory&CPq_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FHarvard%2Farchive_morph&up=CPq&JavaScript=enabled Structural MRI] and [https://portal.nbirn.net:443/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=gotoDirectory&up=7li&7li_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FHarvard%2Farchive_diffusion&JavaScript=enabled DTI data]).
:* Tri worked hard on making sure algorithm works on new high resolution 3T data (available here: [https://portal.nbirn.net/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=moveUpDir&gv7_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FPNL%2F3T_strct_dti_fmri%2Fcase01017&up=gv7&JavaScript=enabled|PNL 3T Data]).
:* Tests have been done also on the spiral diffusion phantom, to make sure diffusion directions and scanner coordinates are handled properly by the algorithm (Figure 6).

{|cellpadding="0"
| style="width:85%" |
; B - Clincal Applications

:* Algorithm was used to trace and analyze anterior limb of the internal capsule on 1.5T data. It generated reacher representation of frontal fiber projections, it also turned out to be more sensitive to group differences in white matter integrity that conventional deterministic tractography (see Figure 6).
:* Algorithm was also used to trace smaller white matter fiber tracts, such as Cingulum, Fornix, Uncinate Fasciculus, Arcuate Fasciculus on 3T "Santa Fe" dataset (http://www.na-mic.org/Wiki/index.php/SanteFe.Tractography.Conference)

; C - References

:* [http://www.na-mic.org/Wiki/index.php/Image:IC_posternew.png Shenton, M.E., Ngo, T., Rosenberger, G., Westin, C.F., Levitt, J.J., McCarley, R.W., Kubicki, M. Study of Thalamo-Cortical White Matter Fiber Tract Projections in Schizophrenia Using Diffusion Stochastic Tractography. Poster presented at the 46th Meeting of the American College of Neuropsychopharmacology, Boca Raton, FL, December 2007.]

== Work in Progress ==

; A - Optimization and Testing of stochastic tractography module :
:* Julien is testing now separate components of work flow, including the masks, and impact of their precision on stochastic output, as well as impact of number of seeding points on tractography results.
:* At the same time, we are testing the module on Max Plank dataset, as well as running it on tractography comparison project dataset.
:* We are discussing possibility of adding project specific functionality to module, such as third ROI to guide tractography, nonlinear registration button for merging fMRI and DTI data.

; B - Related Clinical Projects

| style="width:85%" |

|}

{|cellpadding="0"

| style="width:85%" |

;* Arcuate Fasciculus Extraction Project
We have started the project of investigating Arcuate Fasciculus using Stochastic Tractography (Figure 7). This structure is especially important in both VCFS and schizophrenia, as it connects language related areas (Brocka and Wernicke's), and is involved in language processing quite disturbed in schizophrenia patients. It also can not be reliably traced using deterministic tractography.
:Project involves:
:* Whole brain segmentation, and automatic extraction of regions interconnected by Arcuate Fasciculus (Inferior frontal and Superior Temporal Gyri).
:* White matter segmentation, in order to prevent algorithm from traveling through the ventricles, where diffusivity is high.
:* Non-linear registration of labelmaps to the DTI space.
:* Seeding tracts. We have piloted it using 5000 seeds per voxel, however it is quite time consuming running it on even most powerful computers in the lab, so we have experimented with smaller number of seeds per voxel. We tested 1000 seeds, which gave virtually identical results.
:* Extracting path of interest, and calculating FA along the path for group comparison. Presentation of previous results for 7 schizophrenics and 12 control subjects, can be found here: [[Media:NAMIC_AHM_Arcuate.ppt|Progress Report Presentation]].
:* Results presentation and paper submission. Abstract was submitted and accepted for presentation at World Biological Psychiatry Symposium in Venice, Italy. Paper is in preparation.

| style="width:85%" | [[Image:STArcuate.jpg|thumb|right|200px|<font size=1>Figure 7: The arcuate fasciculus including seed, midpoint and target ROI's.</font>]]

|}

;* Semantic Network Connectivity Project
We use combination of fMRI and DTI data to define and characterize functional and anatomical connectivity within the semantic processing network in schizophrenia.
:Project involves:
:* fMRI data analysis and identification of functional nodes involved in semantic processing in healthy controls and seubjects with schizophrenia
:* Analysis of functional connectivity (using FSL) between nodes of semantic network
:* Whole brain Voxel Based analysis of DTI data in same population
:* Use of stochastic tractography to identify connections between functional nodes
:* Correlational analysis involving anatomical and functional connectivity data.
:* Results presentation and paper submission. Data was presented at HBM conference in 2007, paper has been submitted to HBM.

;* Study of Default Network
We have started collaboration with Department of Neurophysiology Max Planck Institute in Frankfurt (Anna Rotarska contact person). They have dataset containing DTI and resting state fMRI in schizophrenia, and want to use stochastic tractography to measure integrity of anatomical connections within the default network in schizophrenia. Their fMRI data has been co-registered with anatomical scans, and we are putting it into the DTI space. Once this is done, we will start creating tracts connecting fMRI ROIs. This data will be also used to test robustness of our module, since data was collected on a different scanner (3T Philips).

;* Tractography Comparison Project
We are also working on a [http://www.na-mic.org/Wiki/index.php/SanteFe.Tractography.Conference tractography comparison project]dataset, where we apply stochastic tractography to phantom, as well as test dataset.

===Staffing Plan===

* Sylvain and Yogesh are the DBP resources charged with adapting the tools in the NA-MIC Kit to the DBP needs
* Doug Markant, our NAMIC RA has left the lab, and now Doug Terry, is a new NAMIC RA.
* Julien is our new NAMIC software engineer. He is responsible for improving the STM (Stochastic Tractography Module), and making sure software works with STM compliant datasets. [[Link Progress| Progress]]
* Polina is the algorithm core contact
* Brad is the engineering core contact

===Schedule===

* '''10/2007''' - Optimization of Stochastic Tractography algorythm for 1.5T data.
* '''10/2007''' - Algorythm testing on Santa Fe data set and diffusion phantom.
* '''06/2008''' - Optimization of Stochastic Tractography algorythm for 3T data.
* '''11/2008''' - Slicer 3 module prototype using python.
* '''12/2008''' - Slicer 3 module official release
* '''12/2008''' - Documentation and packaging for dissemination.
* '''12/2008''' - Arcuate Fasciculus results.
* '''01/2009''' - Arcuate Fasciuclus first draft of the paper.
* '''05/2009''' - Distortion correction and nonlinear registration added to the module
* '''05/2009''' - Symposium on tractography, including stochastic methods at World Biological Psychiatry Symposium in Florence, Italy.
* '''05/2009''' - Presentation of Arcuate Fasciculus findings at World Biological Psychiatry Symposium in Florence, Italy.

===Team and Institute===
*PI: Marek Kubicki (kubicki at bwh.harvard.edu)
*DBP2 Investigators: Sylvain Bouix, Yogesh Rathi, Julien de Siebenthal
*NA-MIC Engineering Contact: Brad Davis, Kitware
*NA-MIC Algorithms Contact: Polina Gollard, MIT

===Publications===

''In print''

*[http://www.na-mic.org/pages/Special:Publications?text=Kubicki+AND+Westin+AND+DTI&submit=Search&words=all&title=checked&keywords=checked&authors=checked&abstract=checked&sponsors=checked&searchbytag=checked| NA-MIC Publications Database - Clinical Applications]

*[http://www.na-mic.org/pages/Special:Publications?text=Ngo+AND+Golland&submit=Search&words=all&title=checked&keywords=checked&authors=checked&abstract=checked&sponsors=checked&searchbytag=checked| NA-MIC Publications Database - Algorithms Development]

[[Category: Schizophrenia]] [[Category: Diffusion MRI]] [[Category: Segmentation]]

DBP2:Harvard:Brain Segmentation Roadmap

2009-04-15T16:23:43Z

Dougt: /* Work Accomplished */

Back to [[NA-MIC_Internal_Collaborations|NA-MIC Collaborations]], [[DBP2:Harvard|Harvard DBP 2]]
__NOTOC__
=Stochastic Tractography for VCFS=
== Roadmap ==

The main goal of this project is to develop end-to-end application that would be used to characterize anatomical connectivity abnormalities in the brain of patients with velocardiofacial syndrome (VCFS), and to link this information with deficits in schizophrenia. This page describes the technology roadmap for stochastic tractography, using newly acquired 3T data, NAMIC tools and slicer 3.

== Algorithm ==

{|cellpadding="0"

| style="width:85%" |

; A-Description
* Most tractography methods estimate fibers by tracing the maximum direction of diffusion. A limitation of this approach is that, in practice, several factors introduce uncertainty in the tracking procedure, including, noise, splitting and crossing fibers, head motion and image artifacts. To address this uncertainty, stochastic tractography methods have been developed to quantify the uncertainty associated with estimated fibers (Bjornemo et al., 2002). Method uses a propagation model based on stochastics and regularization, which allows paths originating at one point to branch and return a probability distribution of possible paths. The method utilizes principles of a statistical Monte Carlo method called Sequential Importance Sampling and Resampling (SISR). Based on probability functions, using a sequential importance sampling technique ([http://lmi.bwh.harvard.edu/papers/pdfs/2002/bjornemoMICCAI02.pdf Bjornemo et al., 2002]), one can generate thousands of fibers starting in the same point by sequentially drawing random step directions. This gives a very rich model of the fiber distribution, as contrasted with single fibers produced by conventional tractography methods. Moreover, from a large number of sampled paths, probability maps can be generated, providing better estimates of connectivity between several anatomical locations. A comparison of the algorithms can be seen here. (Figure 1)

| style="width:15%" | [[Image:IC_sto_new.png|thumb|right|200px|<font size=1> Figure 1: Comparison of deterministic and stochastic tractography algorithms</font>]]

|}
{|cellpadding="0"

| style="width:85%" |

; B-Possible Applications
* Since diffusion direction uncertainty within the gray matter is quite significant; principal diffusion direction approaches usually do not work for tracking between two gray matter regions. Thus if one requires finding connections between a priori selected anatomical gray matter regions, defined either by anatomical segmentations (in case of using structural ROI data), or functional activations (in case of megring DTI with fMRI), stochastic tractography seems to be the method of choice. Here is an example of this application to anatomical data (Figure 2, left image) and to fMRI data (Figure 2, right image).

* Stochastic Tractography is also comparable, if not better, in defining large white matter fiber bundles, especially those traveling through white matter regions characterized by increased diffusion uncertainty (fiber crossings). Example of such application to internal capsule. (Figure 3)

| style="width:15%" |[[Image:StochasticPic.PNG|thumb|right|200px|<font size=1>Figure 2: Stochastic Tractography of uncinate fasciculis on anatomical data (left) and cingulum bungle on fMRI scan (right)</font>]]
|} [[Image:IC-comp-new.png|thumb|right|200px|<font size=1>Figure 3: Streamline vs. Stochastic Tractography of the Internal Capsule</font>]]
; C-References

* [http://lmi.bwh.harvard.edu/papers/pdfs/2002/bjornemoMICCAI02.pdf Björnemo M, Brun A, Kikinis R, Westin CF. Regularized stochastic white matter tractography using diffusion tensor MRI. In Fifth International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI'02). Tokyo, Japan, 2002;435-442.]
* [http://lmi.bwh.harvard.edu/papers/pdfs/2006/frimanTMI06.pdf Friman, O., Farneback, G., Westin CF. A Bayesian Approach for Stochastic White Matter Tractography. IEEE Transactions on Medical Imaging, Vol 25, No. 8, Aug. 2006]

[[Image:stoch_menu.png|thumb|right|200px|<font size=1> Figure 4: Python Stochastic Tractography GUI </font>]]

==Module==
Can be found in: MODULES > PYTHON MODULES > PYTHON STOCHASTIC TRACTOGRAPHY
;Functionality of Python Stochastic Tractography module in Slicer 3.0
* IO:
Module reads files (DWI and ROIs) in nhdr format.
* Smoothing:
One can smooth the DWI data (only Gausian smoothing is supported at this time). We recommend it if the data is noisy.
* Brain Mask:
The Brain mask defines the volume in which the tensor will be computed and the tracts evaluated.
*Diffusion Tensor:
This step calculates the tensor and can output anisotropy indices (FA/Mode/Trace)
*Tractography:
Parameters that need to be adjusted:
:1. The amount of tracts that will be seeded from each voxel (we recommend between 500 and 1000 tracts, depending on the workstation power- 1000 tracts per voxel seeded within the large ROI for high resolution DWI can take a long time to compute).
:2. Maximum tract length (in mm), this can eliminate long, unwanted tracts if the regions for which connection is measured are located close to each other [[Image:stoch_menu2a.png|thumb|right|200px|<font size=1> Figure 5: Python Stochastic Tractography GUI, part 2 </font>]]
:3. Step Size: distance between each re-estimation of tensors, usually between 0.5 and 1 mm
:4. Stopping criteria. This can be used on the top of WM mask, to terminate tracts (in case they really want to travel through the CSF, for example).
*Connectivity Map:
This step creates output probability maps.
:1. binary: each voxel is counted only once if at least one fiber pass through it
:2. cumulative: tracts are summed by voxel independently
:3. weighted: tracts are summed by voxel depending on their length

Then, probability maps can be saved as ROIs, and either used directly, or thresholded (at certain probability, step claimed by few publications to remove noise) in slicer to mask and compute average FA, Mode, Trace for entire connection. Diffusion indices can be also weighted by the probability of connection for each voxel.

:* Module documentation can be found here:
:**[[Media:IJdata.tar.gz|Training Dataset]]
:**[[Media:PNLStochasictNew.ppt|Training Presentation]]
:**[[Media:Helix.zip|Sample Helix Dataset]]

==Work Accomplished==
[[Image:helix.png|thumb|right|200px|<font size=1>Figure 6: Stochastic Tractography on Phantom</font>]]
; A - Optimization and testing of stochastic tractography algorithm :
:* Original methodological paper, as well as our first attempts to use the algorithm (CC+ and matlab scripts) have been done on old "NAMIC" 1.5T LSDI data ([https://portal.nbirn.net:443/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=gotoDirectory&CPq_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FHarvard%2Farchive_morph&up=CPq&JavaScript=enabled Structural MRI] and [https://portal.nbirn.net:443/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=gotoDirectory&up=7li&7li_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FHarvard%2Farchive_diffusion&JavaScript=enabled DTI data]).
:* Tri worked hard on making sure algorithm works on new high resolution 3T data (available here: [https://portal.nbirn.net/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=moveUpDir&gv7_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FPNL%2F3T_strct_dti_fmri%2Fcase01017&up=gv7&JavaScript=enabled|PNL 3T Data]).
:* Tests have been done also on the spiral diffusion phantom, to make sure diffusion directions and scanner coordinates are handled properly by the algorithm (Figure 6).

{|cellpadding="0"
| style="width:85%" |
; B - Clincal Applications

:* Algorithm was used to trace and analyze anterior limb of the internal capsule on 1.5T data. It generated reacher representation of frontal fiber projections, it also turned out to be more sensitive to group differences in white matter integrity that conventional deterministic tractography (see Figure 6).
:* Algorithm was also used to trace smaller white matter fiber tracts, such as Cingulum, Fornix, Uncinate Fasciculus, Arcuate Fasciculus on 3T "Santa Fe" dataset (http://www.na-mic.org/Wiki/index.php/SanteFe.Tractography.Conference)

; C - References

:* [http://www.na-mic.org/Wiki/index.php/Image:IC_posternew.png Shenton, M.E., Ngo, T., Rosenberger, G., Westin, C.F., Levitt, J.J., McCarley, R.W., Kubicki, M. Study of Thalamo-Cortical White Matter Fiber Tract Projections in Schizophrenia Using Diffusion Stochastic Tractography. Poster presented at the 46th Meeting of the American College of Neuropsychopharmacology, Boca Raton, FL, December 2007.]

== Work in Progress ==

; A - Optimization and Testing of stochastic tractography module :
:* Julien is testing now separate components of work flow, including the masks, and impact of their precision on stochastic output, as well as impact of number of seeding points on tractography results.
:* At the same time, we are testing the module on Max Plank dataset, as well as running it on tractography comparison project dataset.
:* We are discussing possibility of adding project specific functionality to module, such as third ROI to guide tractography, nonlinear registration button for merging fMRI and DTI data.

; B - Related Clinical Projects

| style="width:85%" |

|}

{|cellpadding="0"

| style="width:85%" |

;* Arcuate Fasciculus Extraction Project
We have started the project of investigating Arcuate Fasciculus using Stochastic Tractography (Figure 7). This structure is especially important in both VCFS and schizophrenia, as it connects language related areas (Brocka and Wernicke's), and is involved in language processing quite disturbed in schizophrenia patients. It also can not be reliably traced using deterministic tractography.
:Project involves:
:* Whole brain segmentation, and automatic extraction of regions interconnected by Arcuate Fasciculus (Inferior frontal and Superior Temporal Gyri).
:* White matter segmentation, in order to prevent algorithm from traveling through the ventricles, where diffusivity is high.
:* Non-linear registration of labelmaps to the DTI space.
:* Seeding tracts. We have piloted it using 5000 seeds per voxel, however it is quite time consuming running it on even most powerful computers in the lab, so we have experimented with smaller number of seeds per voxel. We tested 1000 seeds, which gave virtually identical results.
:* Extracting path of interest, and calculating FA along the path for group comparison. Presentation of previous results for 7 schizophrenics and 12 control subjects, can be found here: [[Media:NAMIC_AHM_Arcuate.ppt|Progress Report Presentation]].
:* Results presentation and paper submission. Abstract was submitted and accepted for presentation at World Biological Psychiatry Symposium in Venice, Italy. Paper is in preparation.

| style="width:85%" | [[Image:STArcuate.jpg|thumb|right|200px|<font size=1>Figure 7: The arcuate fasciculus including seed, midpoint and target ROI's.</font>]]

|}
{|cellpadding="0"

| style="width:85%" |

;* Semantic Network Connectivity Project
We use combination of fMRI and DTI data to define and characterize functional and anatomical connectivity within the semantic processing network in schizophrenia.
:Project involves:
:* fMRI data analysis and identification of functional nodes involved in semantic processing in healthy controls and seubjects with schizophrenia
:* Analysis of functional connectivity (using FSL) between nodes of semantic network
:* Whole brain Voxel Based analysis of DTI data in same population
:* Use of stochastic tractography to identify connections between functional nodes
:* Correlational analysis involving anatomical and functional connectivity data.
:* Results presentation and paper submission. Data was presented at HBM conference in 2007, paper has been submitted to HBM.

;* Study of Default Network
We have started collaboration with Department of Neurophysiology Max Planck Institute in Frankfurt (Anna Rotarska contact person). They have dataset containing DTI and resting state fMRI in schizophrenia, and want to use stochastic tractography to measure integrity of anatomical connections within the default network in schizophrenia. Their fMRI data has been co-registered with anatomical scans, and we are putting it into the DTI space. Once this is done, we will start creating tracts connecting fMRI ROIs. This data will be also used to test robustness of our module, since data was collected on a different scanner (3T Philips).

;* Tractography Comparison Project
We are also working on a [http://www.na-mic.org/Wiki/index.php/SanteFe.Tractography.Conference tractography comparison project]dataset, where we apply stochastic tractography to phantom, as well as test dataset.

===Staffing Plan===

* Sylvain and Yogesh are the DBP resources charged with adapting the tools in the NA-MIC Kit to the DBP needs
* Doug Markant, our NAMIC RA has left the lab, and now Doug Terry, is a new NAMIC RA.
* Julien is our new NAMIC software engineer. He is responsible for improving the STM (Stochastic Tractography Module), and making sure software works with STM compliant datasets. [[Link Progress| Progress]]
* Polina is the algorithm core contact
* Brad is the engineering core contact

| style="width:15%" |
|}

===Schedule===

* '''10/2007''' - Optimization of Stochastic Tractography algorythm for 1.5T data.
* '''10/2007''' - Algorythm testing on Santa Fe data set and diffusion phantom.
* '''06/2008''' - Optimization of Stochastic Tractography algorythm for 3T data.
* '''11/2008''' - Slicer 3 module prototype using python.
* '''12/2008''' - Slicer 3 module official release
* '''12/2008''' - Documentation and packaging for dissemination.
* '''12/2008''' - Arcuate Fasciculus results.
* '''01/2009''' - Arcuate Fasciuclus first draft of the paper.
* '''05/2009''' - Distortion correction and nonlinear registration added to the module
* '''05/2009''' - Symposium on tractography, including stochastic methods at World Biological Psychiatry Symposium in Florence, Italy.
* '''05/2009''' - Presentation of Arcuate Fasciculus findings at World Biological Psychiatry Symposium in Florence, Italy.

===Team and Institute===
*PI: Marek Kubicki (kubicki at bwh.harvard.edu)
*DBP2 Investigators: Sylvain Bouix, Yogesh Rathi, Julien de Siebenthal
*NA-MIC Engineering Contact: Brad Davis, Kitware
*NA-MIC Algorithms Contact: Polina Gollard, MIT

===Publications===

''In print''

*[http://www.na-mic.org/pages/Special:Publications?text=Kubicki+AND+Westin+AND+DTI&submit=Search&words=all&title=checked&keywords=checked&authors=checked&abstract=checked&sponsors=checked&searchbytag=checked| NA-MIC Publications Database - Clinical Applications]

*[http://www.na-mic.org/pages/Special:Publications?text=Ngo+AND+Golland&submit=Search&words=all&title=checked&keywords=checked&authors=checked&abstract=checked&sponsors=checked&searchbytag=checked| NA-MIC Publications Database - Algorithms Development]

[[Category: Schizophrenia]] [[Category: Diffusion MRI]] [[Category: Segmentation]]

DBP2:Harvard:Brain Segmentation Roadmap

2009-04-15T16:11:11Z

Dougt: /* Module */

Back to [[NA-MIC_Internal_Collaborations|NA-MIC Collaborations]], [[DBP2:Harvard|Harvard DBP 2]]
__NOTOC__
=Stochastic Tractography for VCFS=
== Roadmap ==

The main goal of this project is to develop end-to-end application that would be used to characterize anatomical connectivity abnormalities in the brain of patients with velocardiofacial syndrome (VCFS), and to link this information with deficits in schizophrenia. This page describes the technology roadmap for stochastic tractography, using newly acquired 3T data, NAMIC tools and slicer 3.

== Algorithm ==

{|cellpadding="0"

| style="width:85%" |

; A-Description
* Most tractography methods estimate fibers by tracing the maximum direction of diffusion. A limitation of this approach is that, in practice, several factors introduce uncertainty in the tracking procedure, including, noise, splitting and crossing fibers, head motion and image artifacts. To address this uncertainty, stochastic tractography methods have been developed to quantify the uncertainty associated with estimated fibers (Bjornemo et al., 2002). Method uses a propagation model based on stochastics and regularization, which allows paths originating at one point to branch and return a probability distribution of possible paths. The method utilizes principles of a statistical Monte Carlo method called Sequential Importance Sampling and Resampling (SISR). Based on probability functions, using a sequential importance sampling technique ([http://lmi.bwh.harvard.edu/papers/pdfs/2002/bjornemoMICCAI02.pdf Bjornemo et al., 2002]), one can generate thousands of fibers starting in the same point by sequentially drawing random step directions. This gives a very rich model of the fiber distribution, as contrasted with single fibers produced by conventional tractography methods. Moreover, from a large number of sampled paths, probability maps can be generated, providing better estimates of connectivity between several anatomical locations. A comparison of the algorithms can be seen here. (Figure 1)

| style="width:15%" | [[Image:IC_sto_new.png|thumb|right|200px|<font size=1> Figure 1: Comparison of deterministic and stochastic tractography algorithms</font>]]

|}
{|cellpadding="0"

| style="width:85%" |

; B-Possible Applications
* Since diffusion direction uncertainty within the gray matter is quite significant; principal diffusion direction approaches usually do not work for tracking between two gray matter regions. Thus if one requires finding connections between a priori selected anatomical gray matter regions, defined either by anatomical segmentations (in case of using structural ROI data), or functional activations (in case of megring DTI with fMRI), stochastic tractography seems to be the method of choice. Here is an example of this application to anatomical data (Figure 2, left image) and to fMRI data (Figure 2, right image).

* Stochastic Tractography is also comparable, if not better, in defining large white matter fiber bundles, especially those traveling through white matter regions characterized by increased diffusion uncertainty (fiber crossings). Example of such application to internal capsule. (Figure 3)

| style="width:15%" |[[Image:StochasticPic.PNG|thumb|right|200px|<font size=1>Figure 2: Stochastic Tractography of uncinate fasciculis on anatomical data (left) and cingulum bungle on fMRI scan (right)</font>]]
|} [[Image:IC-comp-new.png|thumb|right|200px|<font size=1>Figure 3: Streamline vs. Stochastic Tractography of the Internal Capsule</font>]]
; C-References

* [http://lmi.bwh.harvard.edu/papers/pdfs/2002/bjornemoMICCAI02.pdf Björnemo M, Brun A, Kikinis R, Westin CF. Regularized stochastic white matter tractography using diffusion tensor MRI. In Fifth International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI'02). Tokyo, Japan, 2002;435-442.]
* [http://lmi.bwh.harvard.edu/papers/pdfs/2006/frimanTMI06.pdf Friman, O., Farneback, G., Westin CF. A Bayesian Approach for Stochastic White Matter Tractography. IEEE Transactions on Medical Imaging, Vol 25, No. 8, Aug. 2006]

[[Image:stoch_menu.png|thumb|right|200px|<font size=1> Figure 4: Python Stochastic Tractography GUI </font>]]

==Module==
Can be found in: MODULES > PYTHON MODULES > PYTHON STOCHASTIC TRACTOGRAPHY
;Functionality of Python Stochastic Tractography module in Slicer 3.0
* IO:
Module reads files (DWI and ROIs) in nhdr format.
* Smoothing:
One can smooth the DWI data (only Gausian smoothing is supported at this time). We recommend it if the data is noisy.
* Brain Mask:
The Brain mask defines the volume in which the tensor will be computed and the tracts evaluated.
*Diffusion Tensor:
This step calculates the tensor and can output anisotropy indices (FA/Mode/Trace)
*Tractography:
Parameters that need to be adjusted:
:1. The amount of tracts that will be seeded from each voxel (we recommend between 500 and 1000 tracts, depending on the workstation power- 1000 tracts per voxel seeded within the large ROI for high resolution DWI can take a long time to compute).
:2. Maximum tract length (in mm), this can eliminate long, unwanted tracts if the regions for which connection is measured are located close to each other [[Image:stoch_menu2a.png|thumb|right|200px|<font size=1> Figure 5: Python Stochastic Tractography GUI, part 2 </font>]]
:3. Step Size: distance between each re-estimation of tensors, usually between 0.5 and 1 mm
:4. Stopping criteria. This can be used on the top of WM mask, to terminate tracts (in case they really want to travel through the CSF, for example).
*Connectivity Map:
This step creates output probability maps.
:1. binary: each voxel is counted only once if at least one fiber pass through it
:2. cumulative: tracts are summed by voxel independently
:3. weighted: tracts are summed by voxel depending on their length

Then, probability maps can be saved as ROIs, and either used directly, or thresholded (at certain probability, step claimed by few publications to remove noise) in slicer to mask and compute average FA, Mode, Trace for entire connection. Diffusion indices can be also weighted by the probability of connection for each voxel.

:* Module documentation can be found here:
:**[[Media:IJdata.tar.gz|Training Dataset]]
:**[[Media:PNLStochasictNew.ppt|Training Presentation]]
:**[[Media:Helix.zip|Sample Helix Dataset]]

==Work Accomplished==
[[Image:helix.png|thumb|right|200px|<font size=1>Figure 6: Stochastic Tractography on Phantom</font>]]
; A - Optimization and testing of stochastic tractography algorythm :
:* Original methodological paper, as well as our first attempts to use the algorythm (CC+ and matlab scripts) have been done on old "NAMIC" 1.5T LSDI data ([https://portal.nbirn.net:443/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=gotoDirectory&CPq_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FHarvard%2Farchive_morph&up=CPq&JavaScript=enabled Structural MRI] and [https://portal.nbirn.net:443/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=gotoDirectory&up=7li&7li_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FHarvard%2Farchive_diffusion&JavaScript=enabled DTI data]).
:* Tri worked hard on making sure algorythm works on new high resolution 3T data (available here: [https://portal.nbirn.net/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=moveUpDir&gv7_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FPNL%2F3T_strct_dti_fmri%2Fcase01017&up=gv7&JavaScript=enabled|PNL 3T Data]).
:* Tests have been done also on the spiral diffusion phantom, to make sure diffusion directions and scanner coordinates are handled properly by the algorythm (Figure 6).

{|cellpadding="0"
| style="width:85%" |
; B - Clincal Applications

:* Algorythm was used to trace and analyze anterior limb of the internal capsule on 1.5T data. It generated reacher representation of frontal fiber projections, it also turned out to be more sensitive to group differences in white matter integrity that conventional deterministic tractography (see Figure 6).
:* Algorythm was also used to trace smaller white matter fiber tracts, such as Cingulum, Fornix, Uncinate Fasciculus, Arcuate Fasciculus on 3T "Santa Fe" dataset (http://www.na-mic.org/Wiki/index.php/SanteFe.Tractography.Conference)

; C - References

:* [http://www.na-mic.org/Wiki/index.php/Image:IC_posternew.png Shenton, M.E., Ngo, T., Rosenberger, G., Westin, C.F., Levitt, J.J., McCarley, R.W., Kubicki, M. Study of Thalamo-Cortical White Matter Fiber Tract Projections in Schizophrenia Using Diffusion Stochastic Tractography. Poster presented at the 46th Meeting of the American College of Neuropsychopharmacology, Boca Raton, FL, December 2007.]

== Work in Progress ==

; A - Optimization and Testing of stochastic tractography module :
:* Julien is testing now separate components of work flow, including the masks, and impact of their precision on stochastic output, as well as impact of number of seeding points on tractography results.
:* At the same time, we are testing the module on Max Plank dataset, as well as running it on tractography comparison project dataset.
:* We are discussing possibility of adding project specific functionality to module, such as third ROI to guide tractography, nonlinear registration button for merging fMRI and DTI data.

; B - Related Clinical Projects

| style="width:85%" |

|}

{|cellpadding="0"

| style="width:85%" |

;* Arcuate Fasciculus Extraction Project
We have started the project of investigating Arcuate Fasciculus using Stochastic Tractography (Figure 7). This structure is especially important in both VCFS and schizophrenia, as it connects language related areas (Brocka and Wernicke's), and is involved in language processing quite disturbed in schizophrenia patients. It also can not be reliably traced using deterministic tractography.
:Project involves:
:* Whole brain segmentation, and automatic extraction of regions interconnected by Arcuate Fasciculus (Inferior frontal and Superior Temporal Gyri).
:* White matter segmentation, in order to prevent algorithm from traveling through the ventricles, where diffusivity is high.
:* Non-linear registration of labelmaps to the DTI space.
:* Seeding tracts. We have piloted it using 5000 seeds per voxel, however it is quite time consuming running it on even most powerful computers in the lab, so we have experimented with smaller number of seeds per voxel. We tested 1000 seeds, which gave virtually identical results.
:* Extracting path of interest, and calculating FA along the path for group comparison. Presentation of previous results for 7 schizophrenics and 12 control subjects, can be found here: [[Media:NAMIC_AHM_Arcuate.ppt|Progress Report Presentation]].
:* Results presentation and paper submission. Abstract was submitted and accepted for presentation at World Biological Psychiatry Symposium in Venice, Italy. Paper is in preparation.

| style="width:85%" | [[Image:STArcuate.jpg|thumb|right|200px|<font size=1>Figure 7: The arcuate fasciculus including seed, midpoint and target ROI's.</font>]]

|}
{|cellpadding="0"

| style="width:85%" |

;* Semantic Network Connectivity Project
We use combination of fMRI and DTI data to define and characterize functional and anatomical connectivity within the semantic processing network in schizophrenia.
:Project involves:
:* fMRI data analysis and identification of functional nodes involved in semantic processing in healthy controls and seubjects with schizophrenia
:* Analysis of functional connectivity (using FSL) between nodes of semantic network
:* Whole brain Voxel Based analysis of DTI data in same population
:* Use of stochastic tractography to identify connections between functional nodes
:* Correlational analysis involving anatomical and functional connectivity data.
:* Results presentation and paper submission. Data was presented at HBM conference in 2007, paper has been submitted to HBM.

;* Study of Default Network
We have started collaboration with Department of Neurophysiology Max Planck Institute in Frankfurt (Anna Rotarska contact person). They have dataset containing DTI and resting state fMRI in schizophrenia, and want to use stochastic tractography to measure integrity of anatomical connections within the default network in schizophrenia. Their fMRI data has been co-registered with anatomical scans, and we are putting it into the DTI space. Once this is done, we will start creating tracts connecting fMRI ROIs. This data will be also used to test robustness of our module, since data was collected on a different scanner (3T Philips).

;* Tractography Comparison Project
We are also working on a [http://www.na-mic.org/Wiki/index.php/SanteFe.Tractography.Conference tractography comparison project]dataset, where we apply stochastic tractography to phantom, as well as test dataset.

===Staffing Plan===

* Sylvain and Yogesh are the DBP resources charged with adapting the tools in the NA-MIC Kit to the DBP needs
* Doug Markant, our NAMIC RA has left the lab, and now Doug Terry, is a new NAMIC RA.
* Julien is our new NAMIC software engineer. He is responsible for improving the STM (Stochastic Tractography Module), and making sure software works with STM compliant datasets. [[Link Progress| Progress]]
* Polina is the algorithm core contact
* Brad is the engineering core contact

| style="width:15%" |
|}

===Schedule===

* '''10/2007''' - Optimization of Stochastic Tractography algorythm for 1.5T data.
* '''10/2007''' - Algorythm testing on Santa Fe data set and diffusion phantom.
* '''06/2008''' - Optimization of Stochastic Tractography algorythm for 3T data.
* '''11/2008''' - Slicer 3 module prototype using python.
* '''12/2008''' - Slicer 3 module official release
* '''12/2008''' - Documentation and packaging for dissemination.
* '''12/2008''' - Arcuate Fasciculus results.
* '''01/2009''' - Arcuate Fasciuclus first draft of the paper.
* '''05/2009''' - Distortion correction and nonlinear registration added to the module
* '''05/2009''' - Symposium on tractography, including stochastic methods at World Biological Psychiatry Symposium in Florence, Italy.
* '''05/2009''' - Presentation of Arcuate Fasciculus findings at World Biological Psychiatry Symposium in Florence, Italy.

===Team and Institute===
*PI: Marek Kubicki (kubicki at bwh.harvard.edu)
*DBP2 Investigators: Sylvain Bouix, Yogesh Rathi, Julien de Siebenthal
*NA-MIC Engineering Contact: Brad Davis, Kitware
*NA-MIC Algorithms Contact: Polina Gollard, MIT

===Publications===

''In print''

*[http://www.na-mic.org/pages/Special:Publications?text=Kubicki+AND+Westin+AND+DTI&submit=Search&words=all&title=checked&keywords=checked&authors=checked&abstract=checked&sponsors=checked&searchbytag=checked| NA-MIC Publications Database - Clinical Applications]

*[http://www.na-mic.org/pages/Special:Publications?text=Ngo+AND+Golland&submit=Search&words=all&title=checked&keywords=checked&authors=checked&abstract=checked&sponsors=checked&searchbytag=checked| NA-MIC Publications Database - Algorithms Development]

[[Category: Schizophrenia]] [[Category: Diffusion MRI]] [[Category: Segmentation]]

DBP2:Harvard:Brain Segmentation Roadmap

2009-04-15T16:10:30Z

Dougt: /* Algorithm */

Back to [[NA-MIC_Internal_Collaborations|NA-MIC Collaborations]], [[DBP2:Harvard|Harvard DBP 2]]
__NOTOC__
=Stochastic Tractography for VCFS=
== Roadmap ==

The main goal of this project is to develop end-to-end application that would be used to characterize anatomical connectivity abnormalities in the brain of patients with velocardiofacial syndrome (VCFS), and to link this information with deficits in schizophrenia. This page describes the technology roadmap for stochastic tractography, using newly acquired 3T data, NAMIC tools and slicer 3.

== Algorithm ==

{|cellpadding="0"

| style="width:85%" |

; A-Description
* Most tractography methods estimate fibers by tracing the maximum direction of diffusion. A limitation of this approach is that, in practice, several factors introduce uncertainty in the tracking procedure, including, noise, splitting and crossing fibers, head motion and image artifacts. To address this uncertainty, stochastic tractography methods have been developed to quantify the uncertainty associated with estimated fibers (Bjornemo et al., 2002). Method uses a propagation model based on stochastics and regularization, which allows paths originating at one point to branch and return a probability distribution of possible paths. The method utilizes principles of a statistical Monte Carlo method called Sequential Importance Sampling and Resampling (SISR). Based on probability functions, using a sequential importance sampling technique ([http://lmi.bwh.harvard.edu/papers/pdfs/2002/bjornemoMICCAI02.pdf Bjornemo et al., 2002]), one can generate thousands of fibers starting in the same point by sequentially drawing random step directions. This gives a very rich model of the fiber distribution, as contrasted with single fibers produced by conventional tractography methods. Moreover, from a large number of sampled paths, probability maps can be generated, providing better estimates of connectivity between several anatomical locations. A comparison of the algorithms can be seen here. (Figure 1)

| style="width:15%" | [[Image:IC_sto_new.png|thumb|right|200px|<font size=1> Figure 1: Comparison of deterministic and stochastic tractography algorithms</font>]]

|}
{|cellpadding="0"

| style="width:85%" |

; B-Possible Applications
* Since diffusion direction uncertainty within the gray matter is quite significant; principal diffusion direction approaches usually do not work for tracking between two gray matter regions. Thus if one requires finding connections between a priori selected anatomical gray matter regions, defined either by anatomical segmentations (in case of using structural ROI data), or functional activations (in case of megring DTI with fMRI), stochastic tractography seems to be the method of choice. Here is an example of this application to anatomical data (Figure 2, left image) and to fMRI data (Figure 2, right image).

* Stochastic Tractography is also comparable, if not better, in defining large white matter fiber bundles, especially those traveling through white matter regions characterized by increased diffusion uncertainty (fiber crossings). Example of such application to internal capsule. (Figure 3)

| style="width:15%" |[[Image:StochasticPic.PNG|thumb|right|200px|<font size=1>Figure 2: Stochastic Tractography of uncinate fasciculis on anatomical data (left) and cingulum bungle on fMRI scan (right)</font>]]
|} [[Image:IC-comp-new.png|thumb|right|200px|<font size=1>Figure 3: Streamline vs. Stochastic Tractography of the Internal Capsule</font>]]
; C-References

* [http://lmi.bwh.harvard.edu/papers/pdfs/2002/bjornemoMICCAI02.pdf Björnemo M, Brun A, Kikinis R, Westin CF. Regularized stochastic white matter tractography using diffusion tensor MRI. In Fifth International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI'02). Tokyo, Japan, 2002;435-442.]
* [http://lmi.bwh.harvard.edu/papers/pdfs/2006/frimanTMI06.pdf Friman, O., Farneback, G., Westin CF. A Bayesian Approach for Stochastic White Matter Tractography. IEEE Transactions on Medical Imaging, Vol 25, No. 8, Aug. 2006]

[[Image:stoch_menu.png|thumb|right|200px|<font size=1> Figure 4: Python Stochastic Tractography GUI </font>]]

==Module==
Can be found in: MODULES > PYTHON MODULES > PYTHON STOCHASTIC TRACTOGRAPHY
;Functionality of Python Stochastic Tractography module in Slicer 3.0
* IO:
Module reads files (DWI and ROIs) in nhdr format.
* Smoothing:
One can smooth the DWI data (only Gausian smoothing is supported at this time). We recommend it if the data is noisy.
* Brain Mask:
The Brain mask defines the volume in which the tensor will be computed and the tracts evaluated.
*Diffusion Tensor:
This step calculates the tensor and can output anisotropy indices (FA/Mode/Trace)
*Tractography:
Parameters that need to be adjusted:
:1. The amount of tracts that will be seeded from each voxel (we recommend between 500 and 1000 tracts, depending on the workstation power- 1000 tracts per voxel seeded within the large ROI for high resolution DWI can take a long time to compute).
:2. Maximum tract length (in mm), this can eliminate long, unwanted tracts if the regions for which connection is measured are located close to each other [[Image:stoch_menu2a.png|thumb|right|200px|<font size=1> Figure 5: Python Stochastic Tractography GUI, part 2 </font>]]
:3. Step Size: distance between each re-estimation of tensors, usually between 0.5 and 1 mm
:4. Stopping criteria. This can be used on the top of WM mask, to terminate tracts (in case they really want to travel through the CSF, for example).
*Connectivity Map:
This step creates output probability maps.
:1. binary: each voxel is counted only once if at least one fiber pass through it
:2. cumulative: tracts are summed by voxel independently
:3. weighted: tracts are summed by voxel depending on their length

Then, probability maps can be saved as ROIs, and either used directly, or thresholded (at certain probability, step claimed by few publications to remove noise) in slicer to mask and compute average FA, Mode, Trace for entire connection. Diffusion indices can be also weighted by the probability of connection for each voxel.

:* Module documentation can be found here:
:**[[Media:IJdata.tar.gz|Training Dataset]]
:**[[Media:PNLStochasictNew.ppt|Training Presentation]]
:**[[Media:Helix.zip|Sample Helix Dataset]]

==Work Accomplished==
[[Image:helix.png|thumb|right|200px|<font size=1>Figure 6: Stochastic Tractography on Phantom</font>]]
; A - Optimization and testing of stochastic tractography algorythm :
:* Original methodological paper, as well as our first attempts to use the algorythm (CC+ and matlab scripts) have been done on old "NAMIC" 1.5T LSDI data ([https://portal.nbirn.net:443/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=gotoDirectory&CPq_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FHarvard%2Farchive_morph&up=CPq&JavaScript=enabled Structural MRI] and [https://portal.nbirn.net:443/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=gotoDirectory&up=7li&7li_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FHarvard%2Farchive_diffusion&JavaScript=enabled DTI data]).
:* Tri worked hard on making sure algorythm works on new high resolution 3T data (available here: [https://portal.nbirn.net/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=moveUpDir&gv7_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FPNL%2F3T_strct_dti_fmri%2Fcase01017&up=gv7&JavaScript=enabled|PNL 3T Data]).
:* Tests have been done also on the spiral diffusion phantom, to make sure diffusion directions and scanner coordinates are handled properly by the algorythm (Figure 6).

{|cellpadding="0"
| style="width:85%" |
; B - Clincal Applications

:* Algorythm was used to trace and analyze anterior limb of the internal capsule on 1.5T data. It generated reacher representation of frontal fiber projections, it also turned out to be more sensitive to group differences in white matter integrity that conventional deterministic tractography (see Figure 6).
:* Algorythm was also used to trace smaller white matter fiber tracts, such as Cingulum, Fornix, Uncinate Fasciculus, Arcuate Fasciculus on 3T "Santa Fe" dataset (http://www.na-mic.org/Wiki/index.php/SanteFe.Tractography.Conference)

; C - References

:* [http://www.na-mic.org/Wiki/index.php/Image:IC_posternew.png Shenton, M.E., Ngo, T., Rosenberger, G., Westin, C.F., Levitt, J.J., McCarley, R.W., Kubicki, M. Study of Thalamo-Cortical White Matter Fiber Tract Projections in Schizophrenia Using Diffusion Stochastic Tractography. Poster presented at the 46th Meeting of the American College of Neuropsychopharmacology, Boca Raton, FL, December 2007.]

== Work in Progress ==

; A - Optimization and Testing of stochastic tractography module :
:* Julien is testing now separate components of work flow, including the masks, and impact of their precision on stochastic output, as well as impact of number of seeding points on tractography results.
:* At the same time, we are testing the module on Max Plank dataset, as well as running it on tractography comparison project dataset.
:* We are discussing possibility of adding project specific functionality to module, such as third ROI to guide tractography, nonlinear registration button for merging fMRI and DTI data.

; B - Related Clinical Projects

| style="width:85%" |

|}

{|cellpadding="0"

| style="width:85%" |

;* Arcuate Fasciculus Extraction Project
We have started the project of investigating Arcuate Fasciculus using Stochastic Tractography (Figure 7). This structure is especially important in both VCFS and schizophrenia, as it connects language related areas (Brocka and Wernicke's), and is involved in language processing quite disturbed in schizophrenia patients. It also can not be reliably traced using deterministic tractography.
:Project involves:
:* Whole brain segmentation, and automatic extraction of regions interconnected by Arcuate Fasciculus (Inferior frontal and Superior Temporal Gyri).
:* White matter segmentation, in order to prevent algorithm from traveling through the ventricles, where diffusivity is high.
:* Non-linear registration of labelmaps to the DTI space.
:* Seeding tracts. We have piloted it using 5000 seeds per voxel, however it is quite time consuming running it on even most powerful computers in the lab, so we have experimented with smaller number of seeds per voxel. We tested 1000 seeds, which gave virtually identical results.
:* Extracting path of interest, and calculating FA along the path for group comparison. Presentation of previous results for 7 schizophrenics and 12 control subjects, can be found here: [[Media:NAMIC_AHM_Arcuate.ppt|Progress Report Presentation]].
:* Results presentation and paper submission. Abstract was submitted and accepted for presentation at World Biological Psychiatry Symposium in Venice, Italy. Paper is in preparation.

| style="width:85%" | [[Image:STArcuate.jpg|thumb|right|200px|<font size=1>Figure 7: The arcuate fasciculus including seed, midpoint and target ROI's.</font>]]

|}
{|cellpadding="0"

| style="width:85%" |

;* Semantic Network Connectivity Project
We use combination of fMRI and DTI data to define and characterize functional and anatomical connectivity within the semantic processing network in schizophrenia.
:Project involves:
:* fMRI data analysis and identification of functional nodes involved in semantic processing in healthy controls and seubjects with schizophrenia
:* Analysis of functional connectivity (using FSL) between nodes of semantic network
:* Whole brain Voxel Based analysis of DTI data in same population
:* Use of stochastic tractography to identify connections between functional nodes
:* Correlational analysis involving anatomical and functional connectivity data.
:* Results presentation and paper submission. Data was presented at HBM conference in 2007, paper has been submitted to HBM.

;* Study of Default Network
We have started collaboration with Department of Neurophysiology Max Planck Institute in Frankfurt (Anna Rotarska contact person). They have dataset containing DTI and resting state fMRI in schizophrenia, and want to use stochastic tractography to measure integrity of anatomical connections within the default network in schizophrenia. Their fMRI data has been co-registered with anatomical scans, and we are putting it into the DTI space. Once this is done, we will start creating tracts connecting fMRI ROIs. This data will be also used to test robustness of our module, since data was collected on a different scanner (3T Philips).

;* Tractography Comparison Project
We are also working on a [http://www.na-mic.org/Wiki/index.php/SanteFe.Tractography.Conference tractography comparison project]dataset, where we apply stochastic tractography to phantom, as well as test dataset.

===Staffing Plan===

* Sylvain and Yogesh are the DBP resources charged with adapting the tools in the NA-MIC Kit to the DBP needs
* Doug Markant, our NAMIC RA has left the lab, and now Doug Terry, is a new NAMIC RA.
* Julien is our new NAMIC software engineer. He is responsible for improving the STM (Stochastic Tractography Module), and making sure software works with STM compliant datasets. [[Link Progress| Progress]]
* Polina is the algorithm core contact
* Brad is the engineering core contact

| style="width:15%" |
|}

===Schedule===

* '''10/2007''' - Optimization of Stochastic Tractography algorythm for 1.5T data.
* '''10/2007''' - Algorythm testing on Santa Fe data set and diffusion phantom.
* '''06/2008''' - Optimization of Stochastic Tractography algorythm for 3T data.
* '''11/2008''' - Slicer 3 module prototype using python.
* '''12/2008''' - Slicer 3 module official release
* '''12/2008''' - Documentation and packaging for dissemination.
* '''12/2008''' - Arcuate Fasciculus results.
* '''01/2009''' - Arcuate Fasciuclus first draft of the paper.
* '''05/2009''' - Distortion correction and nonlinear registration added to the module
* '''05/2009''' - Symposium on tractography, including stochastic methods at World Biological Psychiatry Symposium in Florence, Italy.
* '''05/2009''' - Presentation of Arcuate Fasciculus findings at World Biological Psychiatry Symposium in Florence, Italy.

===Team and Institute===
*PI: Marek Kubicki (kubicki at bwh.harvard.edu)
*DBP2 Investigators: Sylvain Bouix, Yogesh Rathi, Julien de Siebenthal
*NA-MIC Engineering Contact: Brad Davis, Kitware
*NA-MIC Algorithms Contact: Polina Gollard, MIT

===Publications===

''In print''

*[http://www.na-mic.org/pages/Special:Publications?text=Kubicki+AND+Westin+AND+DTI&submit=Search&words=all&title=checked&keywords=checked&authors=checked&abstract=checked&sponsors=checked&searchbytag=checked| NA-MIC Publications Database - Clinical Applications]

*[http://www.na-mic.org/pages/Special:Publications?text=Ngo+AND+Golland&submit=Search&words=all&title=checked&keywords=checked&authors=checked&abstract=checked&sponsors=checked&searchbytag=checked| NA-MIC Publications Database - Algorithms Development]

[[Category: Schizophrenia]] [[Category: Diffusion MRI]] [[Category: Segmentation]]

DBP2:Harvard:Brain Segmentation Roadmap

2009-04-15T16:09:35Z

Dougt:

Back to [[NA-MIC_Internal_Collaborations|NA-MIC Collaborations]], [[DBP2:Harvard|Harvard DBP 2]]
__NOTOC__
=Stochastic Tractography for VCFS=
== Roadmap ==

The main goal of this project is to develop end-to-end application that would be used to characterize anatomical connectivity abnormalities in the brain of patients with velocardiofacial syndrome (VCFS), and to link this information with deficits in schizophrenia. This page describes the technology roadmap for stochastic tractography, using newly acquired 3T data, NAMIC tools and slicer 3.

== Algorithm ==

{|cellpadding="0"

| style="width:85%" |

; A-Description
* Most tractography methods estimate fibers by tracing the maximum direction of diffusion. A limitation of this approach is that, in practice, several factors introduce uncertainty in the tracking procedure, including, noise, splitting and crossing fibers, head motion and image artifacts. To address this uncertainty, stochastic tractography methods have been developed to quantify the uncertainty associated with estimated fibers (Bjornemo et al., 2002). Method uses a propagation model based on stochastics and regularization, which allows paths originating at one point to branch and return a probability distribution of possible paths. The method utilizes principles of a statistical Monte Carlo method called Sequential Importance Sampling and Resampling (SISR). Based on probability functions, using a sequential importance sampling technique ([http://lmi.bwh.harvard.edu/papers/pdfs/2002/bjornemoMICCAI02.pdf Bjornemo et al., 2002]), one can generate thousands of fibers starting in the same point by sequentially drawing random step directions. This gives a very rich model of the fiber distribution, as contrasted with single fibers produced by conventional tractography methods. Moreover, from a large number of sampled paths, probability maps can be generated, providing better estimates of connectivity between several anatomical locations. A comparison of the algorithms can be seen here. (Figure 1)

| style="width:15%" | [[Image:IC_sto_new.png|thumb|right|200px|<font size=1> Figure 1: Comparison of deterministic and stochastic tractography algorithms</font>]]

|}
{|cellpadding="0"

| style="width:85%" |

; B-Possible Applications
* Since diffusion direction uncertainty within the gray matter is quite significant; principal diffusion direction approaches usually do not work for tracking between two gray matter regions. Thus if one requires finding connections between a priori selected anatomical gray matter regions, defined either by anatomical segmentations (in case of using structural ROI data), or functional activations (in case of megring DTI with fMRI), stochastic tractography seems to be the method of choice. Here is an example of this application to anatomical data (Figure 2, left image) and to fMRI data (Figure 2, right image).

* Stochastic Tractography is also comparable, if not better, in defining large white matter fiber bundles, especially those traveling through white matter regions characterized by increased diffusion uncertainty (fiber crossings). Example of such application to internal capsule. (Figure 3)

| style="width:15%" |[[Image:StochasticPic.PNG|thumb|right|200px|<font size=1>Figure 2: Stochastic Tractography of uncinate fasciculis on anatomical data (left) and cingulum bungle on fMRI scan (right)</font>]]
|} [[Image:IC-comp-new.png|thumb|right|200px|<font size=1>Figure 3: Streamline vs. Stochastic Tractography of the Internal Capsule</font>]]
; C-References

* [http://lmi.bwh.harvard.edu/papers/pdfs/2002/bjornemoMICCAI02.pdf Björnemo M, Brun A, Kikinis R, Westin CF. Regularized stochastic white matter tractography using diffusion tensor MRI. In Fifth International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI'02). Tokyo, Japan, 2002;435-442.]
* [http://lmi.bwh.harvard.edu/papers/pdfs/2006/frimanTMI06.pdf Friman, O., Farneback, G., Westin CF. A Bayesian Approach for Stochastic White Matter Tractography. IEEE Transactions on Medical Imaging, Vol 25, No. 8, Aug. 2006]

[[Image:stoch_menu.png|thumb|right|200px|<font size=1> Figure 4: Python Stochastic Tractography GUI </font>]]
==Module==
Can be found in: MODULES > PYTHON MODULES > PYTHON STOCHASTIC TRACTOGRAPHY
;Functionality of Python Stochastic Tractography module in Slicer 3.0
* IO:
Module reads files (DWI and ROIs) in nhdr format.
* Smoothing:
One can smooth the DWI data (only Gausian smoothing is supported at this time). We recommend it if the data is noisy.
* Brain Mask:
The Brain mask defines the volume in which the tensor will be computed and the tracts evaluated.
*Diffusion Tensor:
This step calculates the tensor and can output anisotropy indices (FA/Mode/Trace)
*Tractography:
Parameters that need to be adjusted:
:1. The amount of tracts that will be seeded from each voxel (we recommend between 500 and 1000 tracts, depending on the workstation power- 1000 tracts per voxel seeded within the large ROI for high resolution DWI can take a long time to compute).
:2. Maximum tract length (in mm), this can eliminate long, unwanted tracts if the regions for which connection is measured are located close to each other [[Image:stoch_menu2a.png|thumb|right|200px|<font size=1> Figure 5: Python Stochastic Tractography GUI, part 2 </font>]]
:3. Step Size: distance between each re-estimation of tensors, usually between 0.5 and 1 mm
:4. Stopping criteria. This can be used on the top of WM mask, to terminate tracts (in case they really want to travel through the CSF, for example).
*Connectivity Map:
This step creates output probability maps.
:1. binary: each voxel is counted only once if at least one fiber pass through it
:2. cumulative: tracts are summed by voxel independently
:3. weighted: tracts are summed by voxel depending on their length

Then, probability maps can be saved as ROIs, and either used directly, or thresholded (at certain probability, step claimed by few publications to remove noise) in slicer to mask and compute average FA, Mode, Trace for entire connection. Diffusion indices can be also weighted by the probability of connection for each voxel.

:* Module documentation can be found here:
:**[[Media:IJdata.tar.gz|Training Dataset]]
:**[[Media:PNLStochasictNew.ppt|Training Presentation]]
:**[[Media:Helix.zip|Sample Helix Dataset]]

==Work Accomplished==
[[Image:helix.png|thumb|right|200px|<font size=1>Figure 6: Stochastic Tractography on Phantom</font>]]
; A - Optimization and testing of stochastic tractography algorythm :
:* Original methodological paper, as well as our first attempts to use the algorythm (CC+ and matlab scripts) have been done on old "NAMIC" 1.5T LSDI data ([https://portal.nbirn.net:443/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=gotoDirectory&CPq_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FHarvard%2Farchive_morph&up=CPq&JavaScript=enabled Structural MRI] and [https://portal.nbirn.net:443/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=gotoDirectory&up=7li&7li_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FHarvard%2Farchive_diffusion&JavaScript=enabled DTI data]).
:* Tri worked hard on making sure algorythm works on new high resolution 3T data (available here: [https://portal.nbirn.net/gridsphere/gridsphere?cid=srbfilebrowser&gs_action=moveUpDir&gv7_dirpath=%2Fhome%2FProjects%2FNAMIC__0003%2FFiles%2FPNL%2F3T_strct_dti_fmri%2Fcase01017&up=gv7&JavaScript=enabled|PNL 3T Data]).
:* Tests have been done also on the spiral diffusion phantom, to make sure diffusion directions and scanner coordinates are handled properly by the algorythm (Figure 6).

{|cellpadding="0"
| style="width:85%" |
; B - Clincal Applications

:* Algorythm was used to trace and analyze anterior limb of the internal capsule on 1.5T data. It generated reacher representation of frontal fiber projections, it also turned out to be more sensitive to group differences in white matter integrity that conventional deterministic tractography (see Figure 6).
:* Algorythm was also used to trace smaller white matter fiber tracts, such as Cingulum, Fornix, Uncinate Fasciculus, Arcuate Fasciculus on 3T "Santa Fe" dataset (http://www.na-mic.org/Wiki/index.php/SanteFe.Tractography.Conference)

; C - References

:* [http://www.na-mic.org/Wiki/index.php/Image:IC_posternew.png Shenton, M.E., Ngo, T., Rosenberger, G., Westin, C.F., Levitt, J.J., McCarley, R.W., Kubicki, M. Study of Thalamo-Cortical White Matter Fiber Tract Projections in Schizophrenia Using Diffusion Stochastic Tractography. Poster presented at the 46th Meeting of the American College of Neuropsychopharmacology, Boca Raton, FL, December 2007.]

== Work in Progress ==

; A - Optimization and Testing of stochastic tractography module :
:* Julien is testing now separate components of work flow, including the masks, and impact of their precision on stochastic output, as well as impact of number of seeding points on tractography results.
:* At the same time, we are testing the module on Max Plank dataset, as well as running it on tractography comparison project dataset.
:* We are discussing possibility of adding project specific functionality to module, such as third ROI to guide tractography, nonlinear registration button for merging fMRI and DTI data.

; B - Related Clinical Projects

| style="width:85%" |

|}

{|cellpadding="0"

| style="width:85%" |

;* Arcuate Fasciculus Extraction Project
We have started the project of investigating Arcuate Fasciculus using Stochastic Tractography (Figure 7). This structure is especially important in both VCFS and schizophrenia, as it connects language related areas (Brocka and Wernicke's), and is involved in language processing quite disturbed in schizophrenia patients. It also can not be reliably traced using deterministic tractography.
:Project involves:
:* Whole brain segmentation, and automatic extraction of regions interconnected by Arcuate Fasciculus (Inferior frontal and Superior Temporal Gyri).
:* White matter segmentation, in order to prevent algorithm from traveling through the ventricles, where diffusivity is high.
:* Non-linear registration of labelmaps to the DTI space.
:* Seeding tracts. We have piloted it using 5000 seeds per voxel, however it is quite time consuming running it on even most powerful computers in the lab, so we have experimented with smaller number of seeds per voxel. We tested 1000 seeds, which gave virtually identical results.
:* Extracting path of interest, and calculating FA along the path for group comparison. Presentation of previous results for 7 schizophrenics and 12 control subjects, can be found here: [[Media:NAMIC_AHM_Arcuate.ppt|Progress Report Presentation]].
:* Results presentation and paper submission. Abstract was submitted and accepted for presentation at World Biological Psychiatry Symposium in Venice, Italy. Paper is in preparation.

| style="width:85%" | [[Image:STArcuate.jpg|thumb|right|200px|<font size=1>Figure 7: The arcuate fasciculus including seed, midpoint and target ROI's.</font>]]

|}
{|cellpadding="0"

| style="width:85%" |

;* Semantic Network Connectivity Project
We use combination of fMRI and DTI data to define and characterize functional and anatomical connectivity within the semantic processing network in schizophrenia.
:Project involves:
:* fMRI data analysis and identification of functional nodes involved in semantic processing in healthy controls and seubjects with schizophrenia
:* Analysis of functional connectivity (using FSL) between nodes of semantic network
:* Whole brain Voxel Based analysis of DTI data in same population
:* Use of stochastic tractography to identify connections between functional nodes
:* Correlational analysis involving anatomical and functional connectivity data.
:* Results presentation and paper submission. Data was presented at HBM conference in 2007, paper has been submitted to HBM.

;* Study of Default Network
We have started collaboration with Department of Neurophysiology Max Planck Institute in Frankfurt (Anna Rotarska contact person). They have dataset containing DTI and resting state fMRI in schizophrenia, and want to use stochastic tractography to measure integrity of anatomical connections within the default network in schizophrenia. Their fMRI data has been co-registered with anatomical scans, and we are putting it into the DTI space. Once this is done, we will start creating tracts connecting fMRI ROIs. This data will be also used to test robustness of our module, since data was collected on a different scanner (3T Philips).

;* Tractography Comparison Project
We are also working on a [http://www.na-mic.org/Wiki/index.php/SanteFe.Tractography.Conference tractography comparison project]dataset, where we apply stochastic tractography to phantom, as well as test dataset.

===Staffing Plan===

* Sylvain and Yogesh are the DBP resources charged with adapting the tools in the NA-MIC Kit to the DBP needs
* Doug Markant, our NAMIC RA has left the lab, and now Doug Terry, is a new NAMIC RA.
* Julien is our new NAMIC software engineer. He is responsible for improving the STM (Stochastic Tractography Module), and making sure software works with STM compliant datasets. [[Link Progress| Progress]]
* Polina is the algorithm core contact
* Brad is the engineering core contact

| style="width:15%" |
|}

===Schedule===

* '''10/2007''' - Optimization of Stochastic Tractography algorythm for 1.5T data.
* '''10/2007''' - Algorythm testing on Santa Fe data set and diffusion phantom.
* '''06/2008''' - Optimization of Stochastic Tractography algorythm for 3T data.
* '''11/2008''' - Slicer 3 module prototype using python.
* '''12/2008''' - Slicer 3 module official release
* '''12/2008''' - Documentation and packaging for dissemination.
* '''12/2008''' - Arcuate Fasciculus results.
* '''01/2009''' - Arcuate Fasciuclus first draft of the paper.
* '''05/2009''' - Distortion correction and nonlinear registration added to the module
* '''05/2009''' - Symposium on tractography, including stochastic methods at World Biological Psychiatry Symposium in Florence, Italy.
* '''05/2009''' - Presentation of Arcuate Fasciculus findings at World Biological Psychiatry Symposium in Florence, Italy.

===Team and Institute===
*PI: Marek Kubicki (kubicki at bwh.harvard.edu)
*DBP2 Investigators: Sylvain Bouix, Yogesh Rathi, Julien de Siebenthal
*NA-MIC Engineering Contact: Brad Davis, Kitware
*NA-MIC Algorithms Contact: Polina Gollard, MIT

===Publications===

''In print''

*[http://www.na-mic.org/pages/Special:Publications?text=Kubicki+AND+Westin+AND+DTI&submit=Search&words=all&title=checked&keywords=checked&authors=checked&abstract=checked&sponsors=checked&searchbytag=checked| NA-MIC Publications Database - Clinical Applications]

*[http://www.na-mic.org/pages/Special:Publications?text=Ngo+AND+Golland&submit=Search&words=all&title=checked&keywords=checked&authors=checked&abstract=checked&sponsors=checked&searchbytag=checked| NA-MIC Publications Database - Algorithms Development]

[[Category: Schizophrenia]] [[Category: Diffusion MRI]] [[Category: Segmentation]]

2009 Winter Project Week Slicer3Functioning

2009-03-24T20:16:12Z

Dougt: /* Key Investigators */

{|
|[[Image:NAMIC-SLC.jpg|thumb|320px|Return to [[2009_Winter_Project_Week|Project Week Main Page]] ]]
|}

__NOTOC__

===Key Investigators===
* BWH: Doug Terry, Marek Kubicki, Sylvain Bouix

<div style="margin: 20px;">

<div style="width: 27%; float: left; padding-right: 3%;">

<h1>Objective</h1>

To determine the basic functionality of Slicer 3.0 by doing day-to-day post-processing tasks in hopes to help identify problems at an early stage of development.

</div>

<div style="width: 27%; float: left; padding-right: 3%;">

<h1>Approach, Plan</h1>

Loading data into Slicer 3.0 and evaluating the ease/efficiency of basic functions (such as drawing, realigning, coregistration, masking, automatic segmentation, deterministic tractography, whole-brain tractography, model making, etc) in a qualitative way.

</div>

<div style="width: 40%; float: left;">

<h1>Progress</h1>

(Feb 24 2009 and following weeks) Doug attended Slicer3 session at 1249 Boylston with several comments and reports about functionality. All bugs have been documented on the Mantis BugTracker.
(1/9/2009) Doug is still in the process of testing basic functionality and should have a report for Steve & Wendy in the following couple weeks.

</div>

<br style="clear: both;" />

</div>

===References===
*

2009 Winter Project Week Slicer3Functioning

2009-01-09T14:13:28Z

Dougt:

{|
|[[Image:NAMIC-SLC.jpg|thumb|320px|Return to [[2009_Winter_Project_Week|Project Week Main Page]] ]]
|}

__NOTOC__

===Key Investigators===
* BWH: Doug Terry, Marek Kubicki, Sylvain Bouix

<div style="margin: 20px;">

<div style="width: 27%; float: left; padding-right: 3%;">

<h1>Objective</h1>

To determine the basic functionality of Slicer 3.0 by doing day-to-day post-processing tasks in hopes to help identify problems at an early stage of development.

</div>

<div style="width: 27%; float: left; padding-right: 3%;">

<h1>Approach, Plan</h1>

Loading data into Slicer 3.0 and evaluating the ease/efficiency of basic functions (such as drawing, realigning, coregistration, masking, automatic segmentation, deterministic tractography, whole-brain tractography, model making, etc) in a qualitative way.

</div>

<div style="width: 40%; float: left;">

<h1>Progress</h1>

(1/9/2009) Doug is still in the process of testing basic functionality and should have a report for Steve & Wendy in the following couple weeks.

</div>

<br style="clear: both;" />

</div>

===References===
*