Electrode Co-Registration

Discussions
1

Dear Brainstorm experts，
Is it possible that registration to MNI space during the coregistration processing can cause stereotactic electrodes to bend, leading to imprecise localization of the electrodes? How do I fix this problem?

Thanks a lot!
Longzhou Xu

2

If you are using a non-linear MNI registration method, the deformations applied to the space between the subject and the template is not uniform. Some areas are indeed going to appear distorted. This is due to the nature of the human brain, which can vary quite a lot in morphology between individuals.

Therefore this is normal that some electrodes that appear nicely aligned in the patients' anatomy appear distorted or with some misaligned contacts in the MNI template.

3

Thank you very much! I have used the DSIstudio to perform the coregistration, this caused electrodes to significant bend. But the MNI coordinate in Brainstorm without showing bend in all 56 subjects. What is the advantage in Brainstorm MNI normalization?

4

The MNI normalization in Brainstorm is performed by SPM:
https://neuroimage.usc.edu/brainstorm/CoordinateSystems

Brainstorm at the moment cannot display bent electrodes. It displays a straight line that best fits the location of all the contacts of this electrode. You would therefore never see a "bent" electrode in Brainstorm, but you need to check the location of each contact with respect to the electrode, as they are probably not well aligned after the MNI normalization.

5

Did you mean that the MNI coordinates obtained from Brainstorm were the results of a straight line fit rather than the origin value from coregistration?

6

The MNI coordinates of the SEEG contacts computed within Brainstorm with a non-linear MNI normalization with SPM or CAT12 are the actual MNI coordinates, there is no interpolation or fit on a straight line that is performed after. In the normalized space (with a non-linear registration), the contacts might not be aligned on a straight line.

However the SEEG electrode shaft is represented as a straight line:

  • When the electrodes are marked in Brainstorm, the line goes from the "tip" to the "skull entry point" defined when marking the electrode.
  • When importing existing coordinates defined from another program, this line is computed as the linear regression of all the contacts of the electrode.

If this is still unclear, please post screen captures to illustrate your questions.