External Email - Use Caution
Hi, I have been using MNEs volume morphing, but I seem to be running into
an issue where when I try to use the nifti output with an ICA program: the
ICA components don't quite "fill" the brain...? I believe I am using python
2.7.15 and mne 0.17.0 to create the files. Thank you for any advice, Lori
External Email - Use Caution
It's possible that there is some bug with the transformations, or the
visualization (we are sorting through some things that could be related in
a volume viz PR <https://github.com/mne-tools/mne-python/pull/6046>
currently). Can you try to replicate this problem using some minimal
example? I would start by seeing if it can be replicated without the
ICA-component step, for example by morphing some activity and visualizing
it to see if it also fails to fill the brain. If this shows the problem
already, you can share the original MRI, destination MRI, original
activation, and code to generate and visualize the morphed activation.
It's best if this can be done with the `sample` dataset (e.g.,
sample->fsaverage), so try to replicate there, too -- it'll make sharing
and testing easier. But I'm guessing it won't show the problem because
sample and fsaverage have the same data orientation, zooms, etc., which
might not be the case for your data.
Eric
External Email - Use Caution
Thank you! I can open the morphed nifti files that come out of MNE-python
in freeview on top of the anatomy and they look reasonable (e.g. fill the
entire anatomy). I then take these nifti files and feed them into an ICA
and the components that come out are what does not quite fill the entire
brainspace. I have done something similar, in a very round-about fashion
with surface morphing (and then projection into the volume) before the
volume morphing was available and did not encounter this issue. One other
note, I seem to remember somewhere in the process an objection that the
units weren't set/or weren't recognized, but it still processed...
Best, Lori
External Email - Use Caution
Apologies, it was an issue with the masking, nothing to do with MNE.
Thanks for working with me to try and figure it out!
Best,
Lori