Model creation throws a warning for 0114_H_PULM_ALGS

[chlookaburra]

Description

I tried to create a solid model using the segmentations for 0114_H_PULM_ALGS. The GUI throws a Warning (see below) and advises me to not mesh the model. The warning is accompanied by red markers that @ktbolt mentioned yesterday. Could you please clarify what should be done here? I think I will get the hang of it after this first pass. I was also thinking that it might be a good idea to log the steps here for future reference.

Reproduction

I am using SimVascular 2025.12.21 on my Ubuntu 24 computer. I opened the 0114_H_PULM_ALGS project in the GUI.

Expected behavior

I was hoping to create a nice model from the segmentations.

Additional context

No response

Code of Conduct

• I agree to follow this project's Code of Conduct and Contributing Guidelines

[chlookaburra]

(Note: I made sure that there are no "stent" segmentations that @ktbolt warned me about.)

I am also attaching another screenshot here to show the three red markers that are in close proximity to each other.

[chlookaburra]

The markers appear to be in between some of the segments in RPA_04_02.

[chlookaburra]

I dug into @ktbolt 's SimVascular/SimVascular issue adventures and discovered that the warning was created here: SimVascular/SimVascular#1154

In the issue: "I've modified the code to display the model even if it is not a valid geometry and display a popup with a warning. The points of the non-manifold triangles are displayed using red markers. This should allow users to determine how to correct the model by modifying paths and segmentations."

Will figure out how to address this next...

[chlookaburra]

@alisonmarsden mentioned that we loop you in too, @numisveinsson . I'm wondering if this is something you would also be interested in investigating. For reference, I need all the pulmonary geometries in the VMR for my UQ work. Unfortunately, the files are very unorganized, so @ktbolt told me that the best way to go about this is to create the solid models from the segmentation again. I need to do this for 17 healthy pulmonary models, 4 pre-op Alagille models, and 2 pre-op Williams models (and possibly for the post-op models but that is a later problem). I'm focusing on the Alagille ones now.

[ktbolt]

@chlookaburra I modified the RPA_04_03.ctgr segmentation a bit (could do a better job of that, a bit too small) and the entire model now builds; centerlines are generated ok it seems.

And contrary to all expectations the model meshes fine

[ktbolt]

None of the pulmonary models are very pleasing.

@numisveinsson It would be very interesting to see if you can make some beauteous pulmonary models using SeqSeg.