External Email - Use Caution
Dear MNE experts,
I have already done a year ago the segmentation and the BEM models in
fiedltrip. Is there a way to import them in MNE without starting from the
segmentation again ?
Thanks
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.nmr.mgh.harvard.edu/pipermail/mne_analysis/attachments/20200427/0c387c0e/attachment.html
External Email - Use Caution
I would redo the segmentation as it will be just a few hours of CPU time
and you're sure not to mess it up with coord frames etc.
my 2c
Alex
External Email - Use Caution
The only problem is that the FreeSurfer skullstripping and other
segmentation steps do not work well (maybe due to a not so good T1) and it
required manual correction. On the other hand, I have the corrected brain
mask. Could be one solution to run FreeSurfer with this brainmask?
Thanks a lot
External Email - Use Caution
You can run FreeSurfer in discrete steps and correct the brainmask.mgz at
the right point in the processing (probably at the end of -autorecon1). See:
https://surfer.nmr.mgh.harvard.edu/fswiki/ReconAllDevTable
See also "going farther back" here:
https://mne.tools/dev/overview/faq.html#my-watershed-bem-meshes-look-incorrect
Alternatively, if you don't want to mess with FreeSurfer, you could try
dealing with the coordinate frames and such. Usually looking at the output
of `plot_alignment` with the subject/src/trans/bem set properly (and also
plot_bem) will tell you if the forward calculation will do what you want it
to do. See for example:
https://mne.tools/dev/auto_tutorials/source-modeling/plot_forward.html
Eric
External Email - Use Caution
Hi Eric,
thanks this help. Only one question for my understanding. *Is* the
brainmask.mgz used by the watershed algorithm at all ?
External Email - Use Caution
It might not be -- it might be recomputed by the `watershed` code. I'm not
sure, though, you'd have to ask the Freesurfer folks how it actually works.
Eric