Add MuscleMap extension (SlicerMuscleMap)

[Eddowesselink]

This pull request adds the MuscleMap extension.

Extension name: MuscleMap
Repository: https://github.com/Eddowesselink/Slicer3D_MuscleMap
Description: Automated whole-body muscle segmentation inside 3D Slicer.
Type: scripted Python extension (no custom C++ libraries).

The extension has been tested locally in Slicer by installing it from source.

[pieper]

Thanks for working on this contribution 👍

Be sure to use the checklist for guidance on how to organize the repo so that it'll work with the rest of the extension infrastructure. From a quick look at your repo you already have most of the info, so it just nees to be reorganized, for example so that some screenshots and paper links show up for people looking for extensions.

https://github.com/Slicer/ExtensionsIndex/blob/main/.github/PULL_REQUEST_TEMPLATE.md

[Eddowesselink]

@pieper, thank you for your reaction. I've adressed the repo following the request template. Could you please let me know if anything not correct?

[pieper]

@Eddowesselink thanks for the update here. I think the thing that would help is to have the checklist here in the PR conversation for reference (i.e. like in this extension PR) for easy reference of which steps are complete.

[pieper]

@lassoan did you have any other comments on this?

[Eddowesselink]

@pieper @lassoan. Thanks for this. Please let me know the next steps. More then willing to copy-paste the checklist into this PR.

[Eddowesselink]

@pieper @lassoan thanks for all the help so far. I've just made and pushed an update so that MuscleMap is somewhat similar to the CADS and TotalSegmentator package (with inputs, outputs, advanced option for force CPU). Can you please tell me how to proceed from here?

[Eddowesselink]

Dear team,
Thank you for the review and helpful feedback. Please let me know the next steps or if there is anything you would like me to address.

[lassoan]

Everything looks good, the module works beautifully. There is just one thing that would be good if you could fix: segments do not have name or terminology. The implementation would be very simple:

• create a ...labels.csv file: that describes the label name, color, terminology for each label value
• load that .csv file into Slicer (it will be loaded as a color table node)
• use that color table node when you load the segmentation result file

[Eddowesselink]

Dear @lassoan, thank you for the feedback. I have now implemented a workaround that pulls the .json file directly from our GitHub repository so that each label is assigned a corresponding color and label name in the Slicer outputs (see figure). As we update our model sequentially with new muscles, this approach will automatically propagate those updates to the Slicer extension outputs.