Hi All,
I'm trying to source localize MEG data in MNE-Python but have run across an issue with a subject's MRI in the freesurfer watershed algorithm. Namely, the outer skull surface ever so slightly overlaps with the inner skull (I've plotted this in freeview for your viewing pleasure). This is a problem as MNE stops with an error in the make_bem_model function when such overlap occurs (understandably).
I have tried deleting the bottom slices of the T1.mgz and rerunning the watershed algorithm but that actually made the overlap worse. I have also tried using the -atlas parameter which made it a bit better but not quite (that's actually what is plotted). What I would really like to do at this point is manually edit the surfaces to avoid the overlap which I think given the location and minuscule nature of the overlap.
Any tips on how to get this done? Or on how to edit the MRI to avoid this overlap to begin with?
Thanks for your help! Keith https://dl.dropboxusercontent.com/u/5190414/skulloverlap.png
[image: Inline image 1]
Keith Doelling Dept of Psychology Center for Neural Science NYU
science.keithdoelling.com