Talairach to vertex

Discussions
,
1

Hi,

I have Talairach coordinates of ECoG electrodes and would like to get for each electrode the closest vertex of my cortex mesh. Context: I would like to use the ECoG signal to seed activity at the corresponding position and forward model that to the scalp level.

Best
Martin

2

Hi Martin,

What is it more precisely that you don't know how to do?

Francois

3

When you say "Talairach" it is a real AC-centered coordinates system but still in patient space, or a type of MNI normalization?
If this is a normalized Talaraich system, I'm sorry you're on your own for the transformation Talairach=>MNI, we dropped completely this outdated referential. But importing MNI coordinates in Brainstorm is easy (but not very accurate).

4

Hi Francois,

I was not sure how to get for a given electrode (in Talairach space) the corresponding vertex of my cortex model.

It is some kind of MNI normalization. I have the Talaraich coordinates of ECoG electrodes projected on a "generic brain derived from MRIs from the McGill Brain Institute".

So, I now took the following approach.

  • Export the MRI:ICMB152 and the cortex_150002V to matlab.

  • Extract for all vertices the MNI coordinates (P_mni= cs_convert(myMRI,'scs','mni', Cortex.Vertices);

  • I multiplied the MNI coordinates by 1000 (I am not quite sure why this is necessary. However, if I do that i get a decent match.

  • Translate the MNI coordinates to Talairach using this approach http://imaging.mrc-cbu.cam.ac.uk/imaging/MniTalairach#Approach_2:_a_non-linear_transform_of_MNI_to_Talairach

  • Compute for each electrode (in Talaraich coordinates) the distance to all vertices (in Talaraich coordinates)

  • Find the vertex with the minimal distance to my electrode position.

Best
martin

5

Looks good. It's a pity you can't simply get the coordinates in native space.
Do you have a MRI or CT scan post implantation for this patient? You could use this to define the positions of the contacts, or at least to check them.
https://neuroimage.usc.edu/brainstorm/Tutorials/ECoG#ECoG_grid:_G
https://neuroimage.usc.edu/brainstorm/Tutorials/Epileptogenicity#Edit_the_contacts_positions

I multiplied the MNI coordinates by 1000 (I am not quite sure why this is necessary. However, if I do that i get a decent match.

In Brainstorm, all the values are saved in international units, so the distances are in meters, while your Talairach coordinates are probably in millimeters.