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

Tissue Classifier tracking example - changed seeding mask to wm only voxels #851

Merged
merged 2 commits into from Feb 10, 2016

Conversation

Projects
None yet
4 participants
@gabknight
Contributor

gabknight commented Jan 29, 2016

Following discussion in issue #831, this PR changes the seeding mask to voxel with 100% wm partial volume fraction. Seeds placed in voxel with a partial volume fraction of gm > 0.5 created unwanted short streamlines.

Thanks @samuelstjean, @arokem

@arokem

This comment has been minimized.

Member

arokem commented Jan 29, 2016

Thanks @gabknight - for the quick turnaround! @samuelstjean - please confirm that this resolves the issue you reported. And thanks for noticing that!

@samuelstjean

This comment has been minimized.

Contributor

samuelstjean commented Feb 3, 2016

There is some other weird thing I wanted to look at, and completely forgot what it was. Will come back to me in a few days hopefully.

@arokem

This comment has been minimized.

Member

arokem commented Feb 3, 2016

In the meanwhile, could you please confirm that this does resolve your previous complaint?

@samuelstjean

This comment has been minimized.

Contributor

samuelstjean commented Feb 3, 2016

Seems good regarding the small fibers, lots of green fibers are gone though.

screenshot from 2016-02-03 14 33 25

All fibers
screenshot from 2016-02-03 14 36 20

As gab suggested in an internal discussion, adding a length threshold to the tracking api would also help, someday. So it does fix the immediate problem, but as some stuff is gone, anyone who knows anatomy wants to pitch in?

@gabknight

This comment has been minimized.

Contributor

gabknight commented Feb 3, 2016

The partial volume fraction maps are not perfect (e.g. there is false gm between the ventricles and the CC), which explains the stopped green streamlines on the first figure.

Streamlines can easily be filtered by their length a posteriori. It won't be hard to included this filtering in the tracking as well..

@Garyfallidis

This comment has been minimized.

Member

Garyfallidis commented Feb 9, 2016

This looks ready to me. Anyone against merging it?

@Garyfallidis

This comment has been minimized.

Member

Garyfallidis commented Feb 10, 2016

Nobody replied so I am assuming this is okay. Thx Gab and Sam reporting it. Any improvements in another PR please.

Garyfallidis added a commit that referenced this pull request Feb 10, 2016

Merge pull request #851 from gabknight/DOC_update_tracking_example
Tissue Classifier tracking example - changed seeding mask to wm only voxels

@Garyfallidis Garyfallidis merged commit 3b8e246 into nipy:master Feb 10, 2016

1 check passed

continuous-integration/travis-ci/pr The Travis CI build passed
Details

@gabknight gabknight deleted the gabknight:DOC_update_tracking_example branch Aug 29, 2017

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment