I’ve got volume source estimates of evoked responses with mne-python. I want to match these source estimates with some fiber tracking results which are in MNI152 space.
In mne-python the source estimates are morphed to fsaverage, which isn’t in mni152 space as I understood.
I found the following: CoordinateSystems - Free Surfer Wiki and read that one can convert a volume from fsaverage to mni152 space with mri_vol2vol. I’m not entirely sure if that’s the right way to go.
Another point that irritates me is that when I save the morphed source estimates as nii-files the results seem to be very blurry and they are also blurry when converting to mni152 space.
I was also wondering if it’s possible to morph the subject specific source estimates directly to a template which is in MNI152 space instead of morphing it to fsaverage and then converting it to mni space.
Here are some possibilities as workarounds (but possibly not the most optimal).
What if you just import MNI152 T1w image into freesurfer and run a recon-all. Make an src from the MNI data at whatever output resolution you would like. Then do a volumetric source morph with the MNI152 set as your src_to (versus fsaverage as subject).
Another option is just to save out the evoked responses as nifti stacks (x/y/z/time) - and just use freesurfer/FSL/afni/SPM to do the morph to MNI152. This would be done outside of the MNE-python framework. We have done this with AFNI and it works fine. Just use your T1 datasets to define the warping to MNI space, then apply non-linear transform to your activation/evoked images.