Further testing/debugging reveals that the error occurs in TotalSegmentatorTool::UpdatePrepare() method at preview->RemoveLabels(preview->GetAllLabelValues()). More precisely, the LabelSetImage::RemoveLabel method has this->InvokeEvent(LabelRemovedEvent(pixelValue)); which causes the exception. This issue is also applicable for Otsu tool (hence updating the title of this task).

Latest testing reveals that the issue is fixed in Ubuntu 22.04.
I think the reason is Qt6 update.

After testing:

1. Contours are still in red. Not black.❌ -> Checklist updated should suffice.
2. Interpolation is updated immediately. No need to reinit anymore. ✔
3. Interpolation is updated immediately. No need to reinit anymore. ✔
4. Dropdown is still in 3-Dimensional, not on Disabled. ❌ -> Checklist updated could suffice as fix.
5. There is no contour. Dropdown on "Disabled" ❌

Pushed new branch to rMITK MITK: feature/T30353-tool-confirm-exit.

Pushed new branch to rMITK MITK: feature/T30391-python312-update.

Pushed new branch to rMITK MITK: feature/T30300-totalseg-fastbox.

Pushed new branch to rMITK MITK: feature/T30357-medsam-doku.

Discussion result - Change checklist

Discussion result - Evaluate if in case of no selection (no labels), the highlight is enough to indicate a "selected group".

Discussion result - Decided on (a)

Pushed new branch to rMITK MITK: feature/T30355-medsam-tool.

Pushed new branch to rMITK MITK: feature/T30353-totalseg-update.

The issue was discovered when working with @s669m for visualization of CNN output data. The reason for this behaviour, as pointed out by @s434n, could be that MITK is unable to properly bin the pixel values for floating point data to create histogram. The hypothesis seems accurate- the same image when converted to int type works as expected in Volume Visualization view.

Deleted branch from rMITK MITK: feature/T30302-python-version-check.

Fixed with D896

SegmentAnyBone is SAM with an extra attention pathway as well, pretrained on MRI dataset.
They have trained the visual transformer vit_t type model and released new pretrained weights. From code perspective, they have modified down SAM code to include attention map and packs with it in the repo.
Salient points:

MedSAM is same old SAM, pretrained using medical image-based dataset.
They have trained the visual transformer vit_b type model and released new pretrained weights. So fundamentally, there is no change in architecture. However, from code perspective, they have modified/trimmed down SAM code to their application and packs with it in the repo.
Salient points:

Ok, I have updated the task description.

Pushed new branch to rMITK MITK: feature/T30171-python-wrapping-new.

Deleted branch from rMITK MITK: feature/T29540-sam-2d-tool.

Deleted branch from rMITK MITK: feature/T29644-sam-doku.

Deleted branch from rMITK MITK: feature/T29560-totalseg-label-update.

Deleted branch from rMITK MITK: feature/T29604-sam-tool-2.

Deleted branch from rMITK MITK: feature/T30215-totalseg-v2-update.

Deleted branch from rMITK MITK: feature/T30253-sam-message-update.

Deleted branch from rMITK MITK: feature/T30256-add-python-version-check.

Pushed new branch to rMITK MITK: feature/T30302-python-version-check.

All pressing issues are now resolved &/or solutions are available. Hence, closing the task.

Basic version checks are introduced via D888. Further refinements in the code will be follow later on.