[Mne_analysis] morph maps, vertex missing/mismatch

[Björn Herrmann]

Dear list,

I was wondering whether someone has advice on a problem I encountered.
I use the morph maps in matlab to morph brain activity from one
individual to another brain.

I calculate the morph maps in the linux shell:
mne_make_morph_maps --from subjID --to fsurfavr

that works all fine, but when I open the morph maps in matlab
(mne_read_morph_map), then in a handful of my subjects, the number of
vertices of the source subject is one to small. The source space
(mne_read_source_spaces) of that individual shows the correct number
of vertices. Also the output in the linux shell while the morph map is
calculated shows the right number of vertices.

When I morph the individuals brain (subjID above) to a different brain
(not the freesurfer standard brain) then it might work properly with
the correct number of vertices for this subject in the morph map
matrix, but not for others. It seems to be a specific interaction
between source and target brain whether one vertex (so far it was
always just one) cannot not be properly morphed and just gets dropped.

Has someone encountered this problem before and is there a way to
solve it. I currently cannot morph a handful of participants to a
common brain space.

Thank you in advance for any input you may have,
Cheers,

Bjorn



----------
Björn Herrmann

Brain and Mind Institute
The University of Western Ontario
London, Ontario N6A 5B7
Canada

email: herrmann.b at gmail.com
web: www.bjoernherrmann.com