Difference between revisions of "DBP MGH Head and Neck Cancer 2014"

From NAMIC Wiki
Jump to: navigation, search
 
(5 intermediate revisions by 2 users not shown)
Line 3: Line 3:
 
*Goal: Get together with your partners from algorithm and engineering to make plans for the upcoming year.
 
*Goal: Get together with your partners from algorithm and engineering to make plans for the upcoming year.
 
*DBP PI: Greg Sharp
 
*DBP PI: Greg Sharp
**Plan for year 4
+
**Review of year 4 goals
 
*** Hybrid interactive-automatic registration
 
*** Hybrid interactive-automatic registration
**** The actors should have listerners, which recive motion events
+
**** Demonstrated at AAPM 2014
**** Some work is required to give feedback to the algorithm
+
**** Current version uses quasi-static dynamics
*** Fully manual module method could also be in the works
+
**** Future versions can use continuous dynamics, penalties can changed during optimization, can wrap any registration algorithm
 +
**** Collaboration with NAC effort
 
*** Atlas-based segmentation
 
*** Atlas-based segmentation
**** Algorithm work is done, needs engineering
+
**** Released Dec 2013 as standalone
 +
**** Some speed improvements are needed
 
*** DICOM-RT improvements
 
*** DICOM-RT improvements
**** Patient hierarchy for keeping track of patient metadata
+
**** Released Nov 2013 as part of SlicerRT
**** New structure set means it will be possible to use the full set of slicer functionality
+
**** Still needs some bug fixes
**** What science will be enabled
 
 
*** DIR validation module
 
*** DIR validation module
**** SIFT method
+
**** Scripted module in development
 +
**** Vector field
 +
** Plans for remaining 6 months
 +
*** Finish up DICOM-RT and DIR validation projects
 +
*** Documentation, and tutorials
 +
** Plans beyond 6 months
 +
*** Funded grant through Dec 2014
 +
*** Submission of grant with MSKCC
 +
*** Look at R01 or BRP funding
 
*Algorithms as extensions in Slicer: Golland
 
*Algorithms as extensions in Slicer: Golland
 +
** KSlice
 +
*** Chan-Vese optimizer is not well suited for soft tissue
 +
*** We should evaluate with "Fast grow cuts" algorithm
 +
** Patch-based segmentation algorithm development
 +
*** Refinement of boundaries and use of Gaussian process
 
*Engineering workflows in collaboration with DBP: Pieper
 
*Engineering workflows in collaboration with DBP: Pieper
** Update to DICOM infrastructure
+
** New framework for registration
** Improvements to Slicer core
+
*** Manage and control landmarks
** Actors for interactive deformable registration
+
** Integration editor functionality
 +
** Update to DICOM & DICOM-RT infrastructure
 +
*** Improvements to CTK DICOM browser
 
** Extension infrastructure
 
** Extension infrastructure
 +
** Subject hierarchy moved to Slicer core

Latest revision as of 20:41, 8 January 2014

Home < DBP MGH Head and Neck Cancer 2014
Back to AHM_2014 Agenda
  • Time: 1pm-2pm
  • Goal: Get together with your partners from algorithm and engineering to make plans for the upcoming year.
  • DBP PI: Greg Sharp
    • Review of year 4 goals
      • Hybrid interactive-automatic registration
        • Demonstrated at AAPM 2014
        • Current version uses quasi-static dynamics
        • Future versions can use continuous dynamics, penalties can changed during optimization, can wrap any registration algorithm
        • Collaboration with NAC effort
      • Atlas-based segmentation
        • Released Dec 2013 as standalone
        • Some speed improvements are needed
      • DICOM-RT improvements
        • Released Nov 2013 as part of SlicerRT
        • Still needs some bug fixes
      • DIR validation module
        • Scripted module in development
        • Vector field
    • Plans for remaining 6 months
      • Finish up DICOM-RT and DIR validation projects
      • Documentation, and tutorials
    • Plans beyond 6 months
      • Funded grant through Dec 2014
      • Submission of grant with MSKCC
      • Look at R01 or BRP funding
  • Algorithms as extensions in Slicer: Golland
    • KSlice
      • Chan-Vese optimizer is not well suited for soft tissue
      • We should evaluate with "Fast grow cuts" algorithm
    • Patch-based segmentation algorithm development
      • Refinement of boundaries and use of Gaussian process
  • Engineering workflows in collaboration with DBP: Pieper
    • New framework for registration
      • Manage and control landmarks
    • Integration editor functionality
    • Update to DICOM & DICOM-RT infrastructure
      • Improvements to CTK DICOM browser
    • Extension infrastructure
    • Subject hierarchy moved to Slicer core