I'm working on importing sEEG electrode positions from Curry / RAVE into the Brainstorm, and running into alignment issues despite several attempts. Here’s what I’ve done so far:
Coregistered the CT to the MRI and it looked good.
I imported the .pom file (attached) from Curry into Brainstorm, but the electrodes appeared far outside the brain (screenshot attached).
Figured .pom coordinates were in CT space, so I applied the same transformation matrices used during the CT-to-MRI coregistration (vox2ras and reg) to transform the coordinates. Loaded the transformed .pom file. This improved the alignment slightly, but they were still not in the correct anatomical location.
I then switched to using RAVE, which provided multiple coordinate options: originalCoordinates, RAS, MNI, and MRIvoxel. Chose the RAS coordinates and created a plain .txt file with:
No header
First column: contact label
Next 3 columns: X, Y, Z coordinates (tab-separated)
In Brainstorm, I tried importing this file using both:
"Import channel file", and
"Add EEG position > Import from file", selecting the format:
"EEG: ASCII: Name, XYZ (tab-separated), scale = 1'
This placed the electrodes closer to the brain, but they still clearly misaligned (I tested with just RAB contacts).
MRI is in the native space (file content pic attached)
Did not try electrode localization in the brainstorm because some contacts have non-uniform spacing.
Please let me know what is it that im doing wrong and how do i fix it.
I converted RAVE T1 coordinates from RAS-LPS - imported into brainstorm as a txt file scale-1.
its still off. its supposed to be where its in in the pic labelled curry. iIm not sure what wrong. MRIs are in the same space.
@ariyas I have a few follow-up questions regarding Curry.
What steps do you follow in Curry to do the localization ? Do you coregister the CT to the MRI first and then localize the contacts on the coregistered CT or do something else.
Can you share the LPS .pom coordinates from Curry ?
