Scout vertices to MNI coordinates (EEG and default anatomy)

Discussions
1

Hello,

I have a head model exported (.mat file) that contains a forward model of 19 channels and 3000 vertices (user defined). I selected 'Schaefer_100_7net' as the atlas, and saved all scouts. Clicking on the scout.mat file vertices3000.Atlas.Scouts results in 1*number double (see scout_vertice_count.jpg below), and clicking on that reveals vertice index numbers (see scouts_vertice_numbers.jpg below)

scout_vertice_count
scout_vertice_count1214×870 139 KB

scouts_vertice_numbers
scouts_vertice_numbers1728×710 83.4 KB

ertice_number.jpg).

  1. Do the vertice index numbers actually correspond to the index of vertices in headmodel.Gain? For example, does 420 in vertice index in scout vertices correspond to the location as in row 420 in headmodel.Gain?

  2. Can we export the SCS/MNI xyz coordinates of each of the 3000 vertices and in each of the vertices in the atlas? I know this is possible in MRI, but is this possible in the EEG head model? Any directions or links related to this will be very helpful, since we are at a loss in exporting location information from our Gain data and atlas data.

Thank you!

2

Yes, the indices that you seen in the Scout.Vertices correspond to the column index in the headmodel.Gain. However, this is the gain (aka forward model). The locations of the vertices are are located in:

  • For surface source spaces: in the Vertices field in the Surface file that was used to create the HeadModel. This Surface file is stored in the field SurfaceFile in the HeadModel file.

  • For volume source space: in the GridLoc field in the HeadModel file.

Note that all the vertices locations are given as [X,Y,Z] in the Subject Coordinate System (SCS) and given in meters:
https://neuroimage.usc.edu/brainstorm/CoordinateSystems#Subject_Coordinate_System_.28SCS_.2F_CTF.29

Locations are already in SCS. If you want to convert them to MNI space coordinates, check this link:
https://neuroimage.usc.edu/brainstorm/CoordinateSystems#Converting_between_coordinate_systems

3

Thank you so much for the detailed explanation! This helped a lot.

One thing I need to make certain is if SCS and MNI coordinates, or the head model that results in 19ch*3000 vertices result in left - right inversion of source points.

With all conditions equal except the number of vertices and the head model being MNI2009a vs. brainstorm's 3000 vertice head model,source localization results of the two head models show left-right inversion when plotted onto 3d space. Has there been a similar issue with this, or is there a process that can incur such an inversion? (Possibly a difference between SCS and MNI left-right axis.)

Thank you!!

4

The two coordinate systems are different: in the way their origins are defined, and the orientations of the X, Y, Z. Both have [mm] as units, but they are defined for different MRIs (SCS for the Subject, and MNI for the ICBM152). This is the reason a conversion between the systems is needed.

https://neuroimage.usc.edu/brainstorm/CoordinateSystems#Subject_Coordinate_System_.28SCS_.2F_CTF.29
https://neuroimage.usc.edu/brainstorm/CoordinateSystems#MNI_coordinates

I am not following the issue. Please provide more information on what the steps you are taking to see such difference.

5

Hello,

We found the problem causing the left-right inversion was due to the difference in channel layout of Brainstorm and the other model we were using. We don't see any issues with Brainstorm now. Thank you so much for your help!