Hello,

I've just started using Freesurfer this week and I'm quite confused about coordinate systems.

Specifically, I have SEEG electrode locations that have been digitized in MRI RAS space, and I'm trying to apply a transform matrix so that these electrodes will be registered with the same brain that I performed recon-all on using Freesurfer, so that I can do further analysis in MNE python.

I understand that voxel space to MRI RAS space transform matrix is stored in vox2ras and the transform for voxel space to Freesurfer's surface RAS is in vox2ras_tkr of T1.mgz in my subject folder. Since the electrodes are in MRI RAS space, I am applying a transform matrix that is a matrix product of inverse(vox2ras) and vox2ras_tkr. However, this doesn't work out in real life, and I end up with electrodes half way outside the head.

Am I using the correct matrices? Also, T1.affine appears similar to vox2ras but very slightly different. Just hoping that there would be an easier way..