AHM2010:Mind - Revision history

Hjbockholt: /* In Progress since last AHM */

2010-01-07T16:10:49Z

In Progress since last AHM

Hjbockholt at 16:08, 7 January 2010

2010-01-07T16:08:20Z

Hjbockholt at 16:07, 7 January 2010

2010-01-07T16:07:12Z

Hjbockholt at 16:06, 7 January 2010

2010-01-07T16:06:39Z

Hjbockholt at 16:02, 7 January 2010

2010-01-07T16:02:42Z

Hjbockholt at 15:58, 7 January 2010

2010-01-07T15:58:12Z

Hjbockholt: /* Detailed Information about the Pipeline */

2010-01-07T15:43:40Z

Detailed Information about the Pipeline

Hjbockholt: /* MRN Roadmap Project */

2010-01-07T15:26:41Z

MRN Roadmap Project

Hjbockholt: /* Detailed Information about the Pipeline */

2010-01-07T00:15:09Z

Detailed Information about the Pipeline

Hjbockholt: /* Overview */

2010-01-06T23:46:03Z

Overview

@@ Line 80: / Line 80: @@
 {|
 |[[Image:CompareViewFlairLesionDiffWholeBrain.png|thumb|333px|CompareViewFlairLesionDiffWholeBrain]]
 |[[Image:CompareViewFlairLesionDiff.png|thumb|333px|CompareViewFlairLesionDiff]]
 |}

@@ Line 14: / Line 14: @@
 ==MRN Roadmap Project==
 {|
@@ Line 24: / Line 35: @@
 |}
-A good examples of the challenges that remains in terms of lesion segmentation were represented at the 2008 MICCAI Conference [http://www.ia.unc.edu/MSseg/ MS Lesion Segmentation Challenge(organized by Warfield and Styner)] in the form of a grand challenge in segmentation. Our participation in this lesion segmentation contest, in which participants were provided training data-sets with two manual labellings, allowed us to experiment with several approaches to lesion segmentation, e.g., EM-Segment (S. Wells), k-means/bayesian (V. Magnotta), and outlier detection (M. Prastawa). Drawing from this experience and a thorough review of the current literature, we have optimized a novel combination of methods that appears to perform well for lesion mapping in NPSLE. Additionally, this approach may generalize to other neurological disorders, such as MS or vascular dementia.
 ==Detailed Information about the Pipeline==

@@ Line 3: / Line 3: @@
 * DBP: H Jeremy Bockholt (PI), Charles Gasparovic(co-PI), Mark Scully(Engineer), The Mind Research Network
 * Core 2: Steve Pieper, Isomics
 * Contact: H. Jeremy Bockholt, jbockholt@mrn.org

@@ Line 5: / Line 5: @@
 * Consultant: Vincent Magnotta, University of Iowa
 * Contact: H. Jeremy Bockholt, jbockholt@mrn.org
 ==MRN Roadmap Project==

@@ Line 4: / Line 4: @@
 * Core 2: Steve Pieper, Isomics
 * Consultant: Vincent Magnotta, University of Iowa
-* Contact: H. Jeremy Bockholt, jbockholt@mrn.org==MRN Roadmap Project==
+* Contact: H. Jeremy Bockholt, jbockholt@mrn.org
 {|

@@ Line 1: / Line 1: @@
 __NOTOC__
-==MRN Roadmap Project==
+==Team==
 {|
@@ Line 42: / Line 46: @@
 **[http://www.nitrc.org/frs/download.php/569/LesionSegmentationTutorialData.tgz tutorial data]
 **[http://www.nitrc.org/frs/download.php/571/Slicer3Training_WhiteMatterLesions_v2.0.ppt.pdf end-user tutorial]
 ==Outreach==

@@ Line 26: / Line 26: @@
 {|
 |[[Image:Scully Figure1.png|thumb|400px|General Lesion Segmentation Protocol]]
-|[[Image:Scully Figure2.jpg|thumb|400px|Feature Subset]]
+|[[Image:Scully Figure2.jpg|thumb|400px|Feature Subset (A) T1w. (B) T2w. (C) FLAIR. (D)
 |}
 {|

@@ Line 5: / Line 5: @@
 |[[Image:Lupus_acquisition.png‎|thumb|280px|Image Acquisition]]
 |[[Image:Intensity_standardize_view.png|thumb|280px|Slicer3 Module Image Intensity Standardization]]
 |[[Image:Predict_lesion_view.png‎|thumb|280px|Slicer3 Module Lesion Classification]]
 |[[Image:LesionInBrain-PW2009.png‎|thumb|280px|Lesion Model inside Brain]]

@@ Line 23: / Line 23: @@
 ***The method makes use of local morphometric features based on multiple MR sequences, including T1-weighted, T2-weighted, and Fluid Attenuated Inversion Recovery. After preprocessing, including co-registration, brain extraction, bias correction, and intensity standardization, 49 features are calculated for each brain voxel based on local morphometry. At each level of segmenta- tion a supervised classifier takes advantage of a different subset of the features to conservatively segment lesion voxels, passing on more difficult voxels to the next classifier. This multi-level approach allows for a fast lesion classification method with tunable trade-offs between sensitivity and specificity producing accuracy comparable to a human rater.
 {|
-|[[Image:Scully Figure1.png|thumb|333px|General Lesion Segmentation Protocol]]
+|[[Image:Scully Figure1.png|thumb|400px|General Lesion Segmentation Protocol]]
-|[[Image:Scully Figure2.jpg|thumb|333px|Feature Subset]]
+|[[Image:Scully Figure2.jpg|thumb|400px|Feature Subset]]
 |}
 {|
-|[[Image:Scully Figure3.jpg|thumb|333px|Predicted Heat Map, Manual Segmentation, Thresholded Heat Map]]
+|[[Image:Scully Figure3.jpg|thumb|400px|Predicted Heat Map, Manual Segmentation, Thresholded Heat Map]]
-|[[Image:Scully Figure4.png|thumb|333px| ROC Curve]]
+|[[Image:Scully Figure4.png|thumb|400px| ROC Curve]]
 |}

@@ Line 14: / Line 14: @@
 * How does the pipeline compare to state of the art?
-**The state of the art in brain lesion classification is a work in progress in NPSLE, as well as, other disorders. There are many approaches that attempt to solve lesion classification in Multiple Sclerosis (where it is much more common to perform neuroimaging studies). Some of these approaches are automated or semi-automated; however, all automatic approaches suffer a lack of ground truth.  It is difficult for human manual raters to agree on fuzzy boundaries across different image contrasts (e.g., T1, T2, FLAIR).
+**Brain lesion classification in NPSLE is currently accomplished manually or by various algorithms developed for other disorders, such as multiple sclerosis (MS).There are many approaches that attempt to solve lesion classification in MS. Some of these approaches are automated or semi-automated; however, all automatic approaches suffer a lack of ground truth.  It is difficult for human manual raters to agree on fuzzy boundaries across different image contrasts (e.g., T1, T2, FLAIR).
-A good example of the challenge that remains in terms of lesion segmentation was well represented at the 2008 MICCAI Conference [http://www.ia.unc.edu/MSseg/ MS Lesion Segmentation Challenge(organized by Warfield and Styner)] this past September in the form of a grand challenge in segmentation.  Several groups, including this DBP, competed in lesion segmentation contest, participants were first provided training data-sets with two manual labellings, their respective approaches were judged in an on-site, real-time competition on additional novel data sets.
+A good examples of the challenges that remains in terms of lesion segmentation were represented at the 2008 MICCAI Conference [http://www.ia.unc.edu/MSseg/ MS Lesion Segmentation Challenge(organized by Warfield and Styner)] in the form of a grand challenge in segmentation. Our participation in this lesion segmentation contest, in which participants were provided training data-sets with two manual labellings, allowed us to experiment with several approaches to lesion segmentation, e.g., EM-Segment (S. Wells), k-means/bayesian (V. Magnotta), and outlier detection (M. Prastawa). Drawing from this experience and a thorough review of the current literature, we have optimized a novel combination of methods that appears to perform well for lesion mapping in NPSLE. Additionally, this approach may generalize to other neurological disorders, such as MS or vascular dementia.
 ==Detailed Information about the Pipeline==