We in NA-MIC, and our collaborating colleagues from BIRN, NAC and UIowa, are in a unique position to make a substantial contribution to the field of knowledge concerning the validation of medical image processing of Diffusion Tensor Image data. Among our faculty are leaders in the field of not only DTI analysis algorithm development, but also of validation and calibration. We seek to use our unique opportunity for multi-site collaboration to advance knowledge in this area for the benefit of clinical and computational scientists.
- To compare and contrast the results of tractography algorithms included in or supported by the NA-MIC Toolkit on a benchmark dataset of Diffusion Tensor Imaging (DTI) data.
- To develop a framework for systematically and statistically comparing and contrasting these outcome measures mapped to specific manuscript preparation.
- To map out appropriate leadership for each of the proposed manuscripts.
- To initiate development of tutorials for tractography algorithms not yet in the NAMIC Training Compendium.
- To propose methods for NA-MIC benchmarks of calibration and validation of tractography algorithms for further discussion at the 2008 AHM.
- October 1-2, 2007, 8:00 AM- 6:00 PM
- The workshop is being held in the Governors Room at the Inn of the Governors, 101 W. Alameda, Sante Fe, NM 87501, 1-800-234-4534, http://www.innofthegovernors.com/
- We have booked a block of rooms for $169.00/night for single or double occupancy September 30- October 3, 2007 (Sunday through Tuesday nights). Our reservation code is "NAMIC" and the deadline for reservations is August 31, 2007. The price includes a full breakfast buffet, wireless internet access in the lobby, meeting room and sleeping rooms at no additional cost.
- Fly into Albuquerque, NM. Information about shuttles from the airport to the hotel can be found here http://www.sandiashuttle.com/.
Many grateful thanks to John Rasure and Debbie Lynch of the MIND Institute for their assistance in making these arrangements.
- To register, add your name to the list of attendees below and make your flight and hotel reservations
- Questions about logistics and the content of the Conference should be addressed to Randy Gollub (rgollub at partners.org).
- This Conference is supported by the National Alliance for Medical Image Computing (NAMIC), funded by the National Institutes of Health through the NIH Roadmap for Medical Research, Grant U54 EB005149. Information on the National Centers for Biomedical Computing can be obtained from http://nihroadmap.nih.gov/bioinformatics.
- Ross Whitaker, PhD
- Guido Gerig, PhD, Utah
- Casey Goodlett, Utah
- Carl-Fredrik Westin, PhD
- Marek Kubicki, MD, PhD
- Sylvain Bouix, PNL
- Randy Gollub, MD, PhD, Harvard Medical School (Department of Psychiatry and Martinos Center, Department of Radiology, Massachussets General Hospital)
- Sonia Pujol, PhD, Harvard Medical School (Surgical Planning Laboratory, Department of Radiology, Brigham and Women's Hospital)
- H. Jeremy Bockholt, The MIND Institute
- John Melonakos, Georgia Tech
- Ron Kikinis, MD, SPL
- Tom Fletcher, Utah
- Anastasia Yendiki, MGH
- Carlo Pierpoli, MD NIH
- Tri Ngo, MIT/BWH
- Marc Niethammer, BWH
- Lauren O'Donnell, BWH
- Vincent Magnotta, Iowa
- Sylvain Gouttard, Utah
Preparation for Workshop -- Important Information for all attendees
Suggested background reading for workshop:
- Guido's 2006 AHM presentation on Validation strategies for DTI analysis
- [If any of you attended the 2005 ISMRM Workshop on Methods for Quantitative Diffusion of MRI of Human Brain and can get us access to the detailed summary, I have seen it and it might be very useful]
- Presentations by and Fitzpatrick from the MICCAI 2003 Tutorial on Validation in Medical Image Processing
- Peruse the webpages associated with the Retrospective Image Registration Evaluation project
- anyone have access to this?
- Sylvain's segmentation comparison manuscript
- CF Westin's experimental study manuscript
Please complete the following items prior to the workshop.
This is hands-on Conference. All participants must come with their own computer loaded with the calibration data and the results of their own algorithm analysis.
Welcome to Santa Fe Dinner
Sunday September 30th, 7:45 PM La Casa Sena, 125 E. Palace, 505-988-9232 (www.lacasasena.com)
October 1, Meeting 8:30 AM - 6 PM
- 7:30- 8:30 AM Enjoy the breakfast buffet at the hotel
- 8:30 - 8:40 AM Goals of Conference Randy Gollub
- 8:40 - 11:45 AM Presentations of results from each of the participants & discussion: Guidelines: Please use this presentation template.
Report-style, no details of methods known from previous Core-1 meetings, report about processing steps, parameters, user-interaction, user time, structure of results. Summary of difficulties and obstacles
- 8:40 Guido Gerig set stage for series of presentations focusing on tractography methods and results
- 8:50 Marek Kubicki Clinical Description of Fiber Bundles and Regions of Interest
- Session 1: 9:10-10:30
- 9:10 Sonia Pujol (BWH) Streamline Tractography in Slicer using ROI and Whole Brain Seeding
- 9:40 Sylvain Gouttard (Utah) FiberViewer
- 10:05 Tri Ngo (MIT/BWH) Stochastic Tractography
- 10:30 -10:45 Coffee Break (place lunch order)
- Session 2: 10:45-12:30
- 10:45 John Melonakos (Georgia Tech) Geodesic Tractography Segmentation
- 11:10 Ross Whitaker/Tom Fletcher (Utah) Volumetric Connectivity
- 11:35 Casey Goodlett (Utah) DTI Atlas
- 12:00 Lunch
- 1:00 Ross Whitaker to kick-off discussion with summarizing ideas & focus
- 1:15 - 2:15 PM What did we learn?
- Problems with data quality (noise, on-scanner up-interpolation, size of datasets, Eddy current distortion, EPI distortion, artifacts etc.
- Problems with ROI definitions, appropriateness of methods given the data.
- Data issues: NRRD header interpretation, output formats, optimal calculation of tensors, others.
- 2:15 - 6:00 PM Where are we going with DTI processing?
- Towards statistical analysis of resulting measures obtained from processing: Approaches, pitfalls (like trying to count streamlines), statistics of tensors, statistics of FA values, functional analysis, multiple comparison correction, others.
- Discuss methods and metrics for contrasting/comparing outcomes of DTI analyses.
- Set detailed agenda for Tuesday
(coffee break somewhere above)
- 7:30 PM Dinner at The Pink Adobe, 406 Old Santa Fe Trail, 505-983-7712.
October 2, 8:30 AM - 5 PM
- 7:30- 8:30 AM Enjoy the breakfast buffet at the hotel
- 8:30 - 8:50 AM Neurosurgical applications of tractography: Lauren O'Donnell (BHW)
- 8:50 - 9:00 AM Pilot results from the MIND multi-site DTI reliability dataset: Jeremy Bockholt (UNM)
- 9:00 - 10:15 Clarify hypotheses/goals for tractography comparison/reliability study
- 10:15 -10:30 Coffee Break
- 10:30 - 1:00 PM Choice of appropriate data set to use for tractography comparison/reliability study
Selection of tracts and methods for ROI selection/generation, agree upon preprocessing algorithm, statistical methods for comparisons Formulation of action plan and assignment of tasks
See Tractography Workshop Core 1 Action Plan for full details including plans for next full group gathering at upcoming NA-MIC AHM
Return to Contrasting Tractography Project Page