Dissemination:Workshop May 26-27 2005UseCaseDiscussion

From NAMIC Wiki
Revision as of 19:21, 18 December 2006 by Andy (talk | contribs) (Update from Wiki)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search
Home < Dissemination:Workshop May 26-27 2005UseCaseDiscussion

Feedback from Users about clinical use cases, and feature requests for Slicer:

Steve et al, please divide this list into two separate lists- one for Slicer development and one for tutorial/training development (e.g. the functionality is already there, just needs to be taught).

  1. Streamline process of MaskStat/Cast for Scalar Volumes (rewrite VolumeMath)
  2. Compute FA along fiber
  3. Telegraphing markers along orthogonal views (Brains feature)
  4. Analyze step editor (can save set of steps for future use)
  5. Distortion correction of DWI (do it and be able to preview/verify it) - might belong in the pre-slicer space
  6. Segmentation: missing the ability to backup (undo)
  7. Keyboard shortcuts should be made available in tutorial
  8. Synchronization of the three planes when you draw
  9. Lots of interest in incorporating Brains Editor functionality in Slicer which is already being explored as a collaboration (nca@unm.edu, vincent_magnotta@uiowa.edu)
  10. DTIStudio (S. Mori) program at JHU is worth looking at
  11. Unify UI: EMSeg seeding is done using C-LMB, while in most other modules Fiducials feature is used for this (press P).
  12. Fiducial feature appears only in 3D window.
  13. Registration GUI needs to be made more uniform
  14. NIfTI readers needed in ITK (format experts to possibly contact: John Ashburner at Dartmouth)
  15. fmri module: discrepancy between 0 indexed (in the configuration parameters) vs. 1 indexed slices (in slicer)
  16. fmri: can we load statistical maps (tmaps) from SPM and threshold it in Slicer? Note that t values can be positive/negative. Typically colorscales (cool/hot) are used. Slicer is most likely to be used for visualization/registration of these tmaps.
  17. fmri: While displaying a tmap, the t-value should show up in the window, not the scaled value.
  18. fmri: want to be able to click on a blob, and get stats on it (SPM gives max and eigenvectors of cluster).
  19. visualization of fmri: be able to show Brodman areas
  20. visualization: how to visualize aggregate data from multiple subjects
  21. add information about slicer contact info to the tutorial slides

From Dartmouth Users:

Slicer: Suggestions for Features/Functionality

  1. would be to have friendly data output format, e.g., deposit volume measures in an excel spreadsheet with ID
  2. "loop" modules for processing multiple subjects in the same way
  3. Paint function for volumetrics
  4. Lesions to normal tissue, visualizations that would be helpful, tools used in structure tracing, features from other programs that are desireable.
  5. group data?
  6. more streamlined as far as memory and processing to prevent crashes
  7. the website specifies the minimum requirements for the computer system, but we would like you to specify "recommended" specs as well
  8. time series analysis - SPM & Slicer
  9. shrink wrap (from Alice)
    • [John MacDonald prefers brush system below to this feature - it doesn't work well]
  10. macros
  11. scripting language
  12. group data?
  13. header checking (automatic parameters)
  14. coronal scrolling
  15. output animations as mp4, avi
  16. option to browse and mark all 3 planes to update the others OR lock any one of the planes
  17. T1 and T2 on the same time (2 registers) -- to get one volume
  18. bump margins -- automatically changes the overlap for structures next to each other
  19. Add manual "paint-brush" editor, with
    • Various sized "brushes" (typically circles, smallest is 1-pixel); these are bi-modal: default mode paints, alt-mode erases. Brush size selected by function or number keys.
    • Brushes should be adjustably translucent; painted color should also be adjustably translucent
    • Modal paint-to-ROI, paint over ROI switch [probably multi paint overlays required to avoid conflicts; eg, if I start tracing lesions I create a lesion overlay space that does not conflict with other overlays; but another time I access an automated model overlay to edit]
    • Modal intensity-capture: a) capture all under brush, b) capture only between threshhold levels
    • Auto-smoothing command key/button
    • Command to auto-close current ROI color, if fully encircled
    • 3-D model build option for visualisation
    • Calculations of ROI areas, dimensions, overall volumes step in tabed text format.
    • Fine brush doubles as drawing tool???
    • Hide/show toggle
    • Requires _large_ interpolated views
    • Ideally, user-definable presets of color groups with names, intensity approximations (later ability to correlate with other objects?).
    • Output options: analyze file if bitmapped overlays, open-GL views and/or quicktime movie, built-in screen/window captures?
    • Outer fringe stuff
      1. ability to segment everyother slice and iterpolate slices in between?
    • Questions:
      1. Is there a need to differentiate a "mask" from an ROI?
  20. Is there a way to do skull-stripping in Slicer at this point? We probably need in the future.
  21. Could use function that would normalize image intensity ranges in irregular scans (esp. FLAIRS).
  22. Would like to be able to place "marks" and see them in all three views
  23. We presently have a realignment phase where several alignments are made (using Brains); would like to be able to do this in slicer with precision, and set origin point at the same time. Output as analyze files.

Dissemination Event: Likes and Improvements/Additions

Likes
  1. We thought there would be problems with the different ability levels but they made it really easy for everyone to participate at an advanced level.
  2. RAs from Marty Shenton's lab were very helpful in getting everyone caught up/solving problems
  3. The slides were extremely helpful, especially with the screen captures so that everyone could see what the menus look like, etc.
Suggestions
  1. Could have been more clear about what we actually needed for the session -- it took 2-3 hours over the workshop that people were getting data, that could have been used for the actual workshop. Need very detailed instructions as to where to put things, in what folder, etc.
  2. Would have been nice to have a smaller group -- especially in the future for more advanced workshops, and maybe the advanced workshops could focus on one area (fMRI vs DTI vs structural)
  3. The feature about bringing your problems and data is desireable, perhaps data with problems so everyone can work through the problem-solving
  4. "Behind-the-scenes" workshop (soon) to understand what is happening, why certain parameters are chosen