Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Skip groom results in points throughout the volume in feature_map_example in Studio #1318

Closed
patkins opened this issue Jun 16, 2021 · 12 comments · Fixed by #1331
Closed

Skip groom results in points throughout the volume in feature_map_example in Studio #1318

patkins opened this issue Jun 16, 2021 · 12 comments · Fixed by #1331
Assignees

Comments

@patkins
Copy link

patkins commented Jun 16, 2021

If the user selects Skip Groom and then Run Optimization, the optimization will place particles throughout the volume instead of on the surface. This occurs even if Run Groom has already been performed.
image
#1273 (the user should probably Run Groom instead of Skip Groom, but I wanted to point this out anyways)

@sheryjoe sheryjoe added this to the 6.1 Release (June 2021) milestone Jun 16, 2021
@akenmorris
Copy link
Contributor

Yeah I guess this goes back to the optimizer input verification issue. If you give the optimizer segmentations instead of distance transforms, it’s going to go haywire.

The “skip grooming” step allows you to import and use distance transforms directly. But that’s not obvious to users, nor are there any checks.

@sheryjoe
Copy link
Contributor

@patkins please mention the issue (and other issues that you submitted) in the relevant cell in the release testing table. thanks

@akenmorris akenmorris self-assigned this Jun 16, 2021
@patkins
Copy link
Author

patkins commented Jun 16, 2021

@sheryjoe I am unable to edit the release testing table, sorry.

@akenmorris
Copy link
Contributor

@sheryjoe , do you have any ideas on how to definitively determine if an image volume is a distance transform or not?

@akenmorris
Copy link
Contributor

@sheryjoe , @patkins , do you think it would be sufficient for now to add a Question Dialog asking the user to make sure their inputs are already distance transforms if they click "skip grooming" for image based domains?

@akenmorris
Copy link
Contributor

Something like this?

image

@akenmorris
Copy link
Contributor

This will display if any of the domains (e.g. multiple domains) are image type.

@patkins
Copy link
Author

patkins commented Jun 17, 2021

@akenmorris I think that's great!

@sheryjoe
Copy link
Contributor

@sheryjoe , do you have any ideas on how to definitively determine if an image volume is a distance transform or not?

@akenmorris it should have both +ve and -ve values and its "zero" level set should be a nonempty mesh, a binary segmentation and a regular image (CT/MRI) should fail this test.

@akenmorris
Copy link
Contributor

Ok, that's a very computationally expensive check. We might consider adding this as part of the optimizer verification check on input, but I don't think we should add this to studio for "skip grooming".

@sheryjoe
Copy link
Contributor

Is checking the min and max of an image expensive? I think VTK isosurface filter throws an error if no surface found. How expensive is this?

@akenmorris
Copy link
Contributor

Well, within Studio for this particular case, I think we are much better off just asking the question. Assuming the user really does have lots of large distance transforms, we will waste a lot of time simply reading them from disk.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging a pull request may close this issue.

3 participants