MNI coordinates of vertices (Colin27 Brain)

Discussions
1

Dear all,
I am working on a combined EEG-fMRI dataset and I would like to extract signal time courses from fMRI-based scouts. For this, I would like to place scouts according to MNI coordinates obtained in my fMRI data analysis. I have the feeling that, in theory, this should be doable in Brainstorm, since the software also returns the MNI-coordinates of my manually placed scouts when I work in the MNI space (using the Colin27 template).

After browsing the support forum, I found this thread (http://neuroimage.usc.edu/forums/showthread.php?1526-Placing-Scouts) which already helped me a bit. However, to really place my seeds according to coordinates, I need to know the MNI coordinates of a vertex in the Colin27 brain. Is there any function or list which relates the 15002 vertices of the template and the MNI coordinate system? I couldn’t find it so far…

Thanks a lot!
Sebastian

2

Hi Sebastian,

Please refer to this tutorial page to start with:
http://neuroimage.usc.edu/brainstorm/CoordinateSystems, section “MNI coordinates”.
You will find at the bottom of the page instructions on how to convert lists of points between coordinate systems.

Note that there is no function doing the transformation between the original subject space and the MNI stereotaxic space in Brainstorm.
What you are trying to do is: importing a 3D mask obtained from your fMRI analysis, and use it in the Colin27 brain (subject that uses the default anatomy) as scouts to extract the activity of your significant fMRI regions?

You can refer to the following posts, and maybe try to contact their authors:
http://neuroimage.usc.edu/forums/showthread.php?1560-Integration-between-EEG-and-FMRI-data
http://neuroimage.usc.edu/forums/showthread.php?1881-fMRI-constrained-EEG-source-analysis-by-Brainstorm

Note that if you have the real anatomy of your subjects, you should use it to estimate the source activity, instead of a template brain that may have a completely different shape.
Then instead of going by MNI coordinates, you could try to use anatomical structures. The spatial resolution we have in the EEG/MEG source modeling is definitely not at the voxel level… you could tolerate for the time-resolved part your analysis to work only at the level of the anatomical region, not one precise set of MNI coordinates. The individual surface-based atlases obtained with FreeSurfer can help you with this.
http://neuroimage.usc.edu/brainstorm/Tutorials/LabelFreeSurfer

Let me know if you need more help.
Francois

3

Dear Francois,

thanks for your help. I will work through the links you suggested and come back to you if I have more questions.

Best,
Sebastian

4

FYI, there is now the possibility to compute MNI coordinates for individual subjects:
http://neuroimage.usc.edu/brainstorm/News#MNI_coordinates