Difference between revisions of "2008 Winter Project Week:Lesions"

From NAMIC Wiki
Jump to: navigation, search
Line 25: Line 25:
 
<h1>Approach, Plan </h1>
 
<h1>Approach, Plan </h1>
 
Fully analyze 3 lupus subjects and 3 matched healthy normal volunteer data-sets collected.
 
Fully analyze 3 lupus subjects and 3 matched healthy normal volunteer data-sets collected.
 
+
Definition of several stages: a)
 +
Criteria and definition of what does an expert define and perceive as a
 +
lesion (like rules on multi-contrast MRI), creation of a
 +
document/catalogue that guides manual segmentation but also algorithm
 +
developers. b) Specification of advanced software for user-guided
 +
efficient segmentation of lesions (e.g. 3D level set evolution etc.,
 +
definition of MRI contrast to use to do that, etc.). c) Towards
 +
approaches towards fully automatic lesion segmentation using
 +
multi-contrast MRI., d) Validateion/test framework for comparison of
 +
user-defined lesion patterns to automated/automatic lesion segmentation,
 +
definition of comparison metrics.
 
</div>
 
</div>
  

Revision as of 18:50, 17 December 2007

Home < 2008 Winter Project Week:Lesions
Lupus.png


Key Investigators

  • MIND: Jeremy Bockolt, Mark Scully
  • BWH: Sonia Pujol
  • Utah: Marcel Prastawa
  • Iowa: Vincent Magnotta
  • Kitware: Brad Davis

Objective

Our objective is to make progress on the roadmap initiative of creating an end to end solution that provides a tutorial for NA-MIC kit users to perform lesion analyses in Slicer3.

Approach, Plan

Fully analyze 3 lupus subjects and 3 matched healthy normal volunteer data-sets collected. Definition of several stages: a) Criteria and definition of what does an expert define and perceive as a lesion (like rules on multi-contrast MRI), creation of a document/catalogue that guides manual segmentation but also algorithm developers. b) Specification of advanced software for user-guided efficient segmentation of lesions (e.g. 3D level set evolution etc., definition of MRI contrast to use to do that, etc.). c) Towards approaches towards fully automatic lesion segmentation using multi-contrast MRI., d) Validateion/test framework for comparison of user-defined lesion patterns to automated/automatic lesion segmentation, definition of comparison metrics.

Progress

June 2007 Project Week

First meeting as a new DBP in lupus. We received hands on training on ITK/VTK, as well as, EM Segment. We processed lupus case using EM Segment.

Jan 2007 Project Half Week

Jeremy attended this meeting in order to learn about NA-MIC



References