I'm working on a source analysis project and trying to retrieve time-series values from an Atlas ROI.
I followed all the instructions on the website, but I can't find the "time series.mat" option among all the Scouts. Can you help me?
Furthermore, I would like to implement a DWI file (.nii), WM fibers files (.trk), and a lesion mapping file (.nii); since all lesions come from a FLAIR scan, I think it might be useful to upload the FLAIR (.nii) scan too, and overlap all images to have all in the same space.
If you want to save the scout time series in a matrix file, you can drag-and-drop the Source file in the Process1 tab and use the process Extract > Scout time series. There is an example of using the process Extract > Scout time series in this tutorial: https://neuroimage.usc.edu/brainstorm/seeg/SeizureFingerprinting#Source_space
What I was looking for is a .mat matrix containing all brain vertices associated with each Scout region of a specific atlas, so that I can identify the actual brain vertices included within each atlas ROI.
I need this information to verify the accuracy of the regions, because I want to see whether all the regions in the selected atlas actually encompass all the brain's vertices.
Thanks for the clarification.
This information is not in the resulting time series file, but in the default cortex file for the Subject, that is the place where the Atlases are stored. In the surface file there is the field Vertices in Surface.Atlas(AtlasIndex).Scouts(ScoutIndex).Vertices that contains the indices of the vertices (Surface.Vertices) for that given ScoutIndex in a given AtlasIndex for the surface
I’m not finding the folder you mention. Where is supposed to be it? Inside each ant subject folder? Cause the generic default subject dosne’t contain that information.
Now, where the vertices exactly are? I would like to know which vertices are in the each Atlas ROI not just the vertex coordinates (in the vertices.mat matrix file I suppose).
I'm encountering this problem: a fter reconstructing the EEG in Source, I load an atlas (attached) with 100 parcels, but some subjects have 102 parcels (attached) and others have 100.
There are few inconsistencies in the way you are estimating sources and how you want to extract the time series.
From the screenshots, we can deduce that sources were computer on the cortical surface
However, the anatomical parcellation (posted .nii file) that you are using to create the scouts is a volume parcellation.
Thus, when creating the surface scouts, there are an not-optimal projection from volume to surface. This projection is not recommended. There is not need to do this projection, there is already a cortical surface version of the Schaefer atlases. Projection from volume to surface leads to an incorrect definition of surface scouts. Check this posts:
All this means, you have to compute the Schaefer cortical parcellations when segmenting the MRI.
Then use these surface atlases to extract time series for surface scouts from surface-space sources.
In an additional note, rather than importing the provided .nii file you could get the Schaefer anatomical parcellations in the MNI space directly in Brainstorm, in this way the colors and labels are correct. But do not project them into the surface, as it was discussed above.
Thank you for your valuable comments. I did everything you told me to do, but some entries show 102 parcels instead of 100 (the correct number). I noticed that some entries in the list are marked with “_net_02” and show 102 parcels, while others are marked with “_net” and show 100 parcels.
The issue seems to be that you are importing multiple atlases with the same name, thus Brainstorm automatically adds the _02 to have unique names. You may want to start clean, delete the previously imported atlas, and start again.
I deleted all other stalases, leaving that one imported suing brainstorm,
but I still have issues with parcels.
Try to run all files at once, I have these not highlighted parcels which are not correct, but if I don't select them I don't get 37, 44, 50 and 28 parcels. I think because some subjects have differ for those, but I'm not figuring out how to solve them
Please double-check the Anatomy atlas that you are using to create these scouts. Since creating them using the provided Schaefer anatomical atlases will set the Scout labels based on the anatomical parcel labels.
That is the reason of the existence of L/R versions of the Scouts. E.g. in the anatomy parcel 37 is completely in the left hemisphere, but when it is projected from anatomy volume to surface, there are some surface vertices in the right hemisphere that are closest to this anatomical parcel (37) than to any anatomical parcel in the right hemisphere. As result two scouts L and R are created.
If you want to use Schaefer atlas in the surface space, compute them for each subject:
You may want to do a sanity check to confirm that the current approach (projecting the anatomical volume parcellation to the cortical surface) is not accurate. Just plot the surface and all the scouts for the projected atlas.
Projected atlas using the approach you are following
) in the Scout tab:
