Engineering:Core 3 Programming Week

Location: 1249 Boylston St, Boston, MA directions

Monday September 26th: Engineering and DBP Discussion

Highlights

Goal: Core 3 to give feedback on how they are using the NAMIC Kit in their daily work.

• noon-1pm: Meet and greet
• 1-2:15pm Feedback from UCI on NAMIC Kit ppt presentation
• 2:15-3:30pm Feedback from Dartmouth on NAMIC Kit ppt presentation
• 3:30-3:50pm break
• 3:50-5:00pm Discussion of features required going forward, Conclusion
• Dinner

Tuesday September 27th Engineering Discussion

9am-noon: Slicer 3.0 Architecure decision making

• Agenda Creation

• Summarize Goals of Slicer 3
  • Serve as Application for Core 3 users of Core 1 Algorithms implemented in ITK
  • Enforce Modularity
  • Broader Development Base (interoperate with different languages, GUI toolkits)

• Technology Discussions
  • Slicer Protocol (Steve)
  • MRML / Coordinate / Publish and Subscribe (Mike)
  • KWWidgets / Widgets / VolView Update (Will)

• Implementation Issues
  • Wrapping ITK Algorithms as Modules
    • Filter Watcher, Build as Standalone Executable or Linked Library
  • UI Definition (GUI and CLI)
    • Command line introspection
    • Progress reporting
  • Process Control (Abort, Sync/Async)
  • Communication
    • Serialization of the Scene, Semantics, Coordinate Systems
    • RPC-like back and forth exchanges (what capabilities are provided by Slicer Base?)
  • Provenance
  • Exporting Scene description to text and binary files

• Timelines and Action Items

• ** By All Hands Meeting
  • • ITK I/O Factory communicating over sockets to slicer 2.x (Steve, Alex)
      • Example executable-based modules:
        • Smooth and image
        • Segmentation
        • Registration
    • TCLAP based XML version of help/interface (Dan, Jim)

• ** At All Hands Meeting

1-2pm

fMRI in slicer review (Wendy, Haiying, Bill, Will, Jim, Steve, Sandy) Background discussion before Polina and Wanmei arrive.

2-5pm: fMRI statistics

Bill, Jim, Will, Sandy, Steve, Wendy, Carsten, Haiying, Polina, Wanmei

• ** *** Current VTK (EMSegmenter) and Matlab (Eric Cosman, Wanmei) based implementations
  • • Migration to ITK and/or improved VTK methods
      • Discuss existing and planned MRF infrastructure [see this discussion on the ITK mailing list]
        • ICM based filters in current ITK
        • Need: MF, Min-Cut/Max-Flow
      • Review existing MI infrastructure and discuss its applicability to activation detection
      • Migrating GLM from GSL to VNL (vtkITK wrapper)
    • Estimates of amount of effort, action items

Wednesday September 28th Engineering Discussion

• 9:15-10:15 Misc:
  • MICCAI Workshop talk (NAMIC Kit: Jim 30 min, NAMIC Overview: Ron 10 min, Slicer: Steve 20 min)
  • BugTracker
  • NIH Forge
• 10:30-noon non-linear ITK registration, widgets