Model Maker does not create model from label map

Slicer built at commit af5f294fb on Linux using GCC 8.2, Python 2.7.15

I just noticed that Model Maker fails to create a model from a label map generated in Simple Growing Segmentation module with a few fiducial points. There’s only one label value.

The error log is here.

Thanks for considering.

Thanks for reporting this.

The error messages from the log:

vtkMRMLColorTableNode::SetColor: ERROR: can't set a colour if not a user defined colour table, reset the type first to User or File

Make me wonder if it’s related to this commit:

It may be related. I can have a look at this later tonight or tomorrow.

I’ve fixed the issue - it was a regression caused by the commit mentioned above. See details here. Nightly builds that are downloaded on Thursday or later should work well.

Model Maker now works as expected.


1 Like

By the way, are you still using Model Maker? Maybe you even use the legacy Editor module? I would recommend to switch to Segment editor, which makes Model Maker module irrelevant, since conversion between various representations happens automatically.

My daily tool in vascular surgery for surgical planning is in fact Volume Rendering. When I need to create an arterial blood flow model from contrast enhanced CT scan, I use Simple Region Growing because I need only one segment and it’s fast.

I have thouroughly explored the new Segmentation Editor and it’s definitely an invaluable tool. Prior painting is rather time consuming however, and this does not suit my workflow.

For instance, with just one fiducial point in the aorta of the CT-Cardio sample dataset, Simple Region Growing does a fabulous job to isolate the arterial lumen as a label map.

Of course, this are many other cases where the new Segment Editor is best suited.