ENH: Add white matter fiber tracking to sEEG tutorial

[alexrockhill]

This adds white matter tracking to the sEEG tutorial so that electrode contacts in white matter can be better interpreted.

[alexrockhill]

Hey @skoudoro and @pjsjongsung, no rush, but when you have a spare few minutes, would you mind reviewing this? Specifically, there's a bit of an issue in the tutorial (tutorials/clinical/20_seeg.py) where the deterministic tracking yields very short, disconnected fibers. I'm wondering if you have any advice on parameters or methods to use to get more helpful fiber tracking. Right now I have it set to find fibers within 50 mm of an sEEG contact in white matter and there's a bunch of disconnected fibers but, ideally, it would be more specific, like 3 mm, and would show the pyramidal tract in this case (LPM 1 is a contact in white matter near left premotor cortex).

[skoudoro]

Hi @alexrockhill, I will look into it later this week. Can you share your peaks (the pam5 contains the peaks). I just want to make sure that they make sense. Not sure if your issue is due to the tractography or the peaks. Last time I encounter such an issue, it was due to the peaks so the reconstruction.

[alexrockhill]

Here's the pam file: https://ohsuitg-my.sharepoint.com/:u:/r/personal/rockhill_ohsu_edu/Documents/pam_example/seeg_sample_pam.pam5?csf=1&web=1&e=u9ZkMk

Thanks for offering to review, really appreciate it.

[alexrockhill]

@larsoner did you do the new release? I'm still getting MNE-misc-data version 0.27 with no diffusion images

[pjsjongsung]

@alexrockhill it says user not in directory and I cannot access the link. Do you have another link you could share or change the permissions of that link?

Edit:
Just got the permission. Thanks!

[alexrockhill]

I added your gmail and UI email. OHSU doesn't allow you to open up permissions unfortunately... I could have shared on Google Drive though... probably should have

[alexrockhill]

https://drive.google.com/file/d/1Fr_VLIoCw-pngGtAAoYB2Wls0obG233-/view?usp=sharing

[alexrockhill]

@larsoner did you do the new release? I'm still getting MNE-misc-data version 0.27 with no diffusion images

Just pinging on this when you have a second @larsoner, thanks

[larsoner]

I cut 0.28 a few weeks ago you need to update mne/datasets/config.py to tell it to use 0.28:

mne-python/mne/datasets/config.py

Line 93 in 5a5b4f1

misc="0.27",

and the hash

mne-python/mne/datasets/config.py

Line 134 in 5a5b4f1

hash="md5:e343d3a00cb49f8a2f719d14f4758afe",

But the version was wrong there (said 0.27 in the dataset itself) so I just cut a 0.29 so use that instead anyway

https://github.com/mne-tools/mne-misc-data/releases/tag/0.29

[alexrockhill]

My bad, I forgot I had to do that, thanks!

[alexrockhill]

Here's the artifact with the rendering of the not-very-good fiber tracking https://output.circle-artifacts.com/output/job/63884b31-afc5-40be-9e27-d1d2264fba57/artifacts/0/html/auto_tutorials/clinical/20_seeg.html#sphx-glr-auto-tutorials-clinical-20-seeg-py

[alexrockhill]

Huh how do you ignore the vulture stuff? It is used... cc @larsoner

[larsoner]

https://github.com/mne-tools/mne-python/blob/main/tools/vulture_allowlist.py

[drammock]

Huh how do you ignore the vulture stuff? It is used...

to find this yourself next time: git grep vulture yields only 10 hits, the last of which is:

pyproject.toml:392:paths = ['mne', 'tools/vulture_allowlist.py']