Dear Brainstorm team,
I would like to know if there is a way to load a functional atlas having “.nii.gz” extension.
Thank you for your help.
Regards,
Aya
Hi Aya,
You can load compressed NIfTI volumes (.nii.gz) in Brainstorm. The problem is more to define what you want to do with it.
If the goal is to use them as regions of interest, you can try loading it from the Scout tab (using the file format “Volume mask or atlas”).
But in general it will give you a very imprecise result.
There are two major problems to consider when trying to use fMRI-based functional atlases for EEG/MEG source analysis:
- The correspondence between volumes (volume masks defined from the fMRI) and surfaces (used for the EEG/MEG source estimation) is difficult to establish.
- You cannot expect the MEG/EEG sources to be located precisely where the fMRI showed there was significant activity because the spatial resolution is a low lower and we are not observing the same effects with the different modalities.
Cheers,
Francois
Thank you Francois for your helpful remarks.
However, I’m very interested in loading the atlas… The problem is that brainstorm gives me the following error statement:
“The selected MRI file does not match the size of the subject’s MRI”
What should I do to solve the problem?
Brainstorm does not do any volume registration at this time.
If you are trying to import two volumes for the same subject (or one structural MRI and one atlas), the two volumes need to be coregistered first (same dimensions, same orientation). You need to do this in an external program.
We are planning to add tools for manipulating MRI volumes more easily, but we currently do not have the necessary resources to develop them.
http://neuroimage.usc.edu/brainstorm/Next#Anatomy
[QUOTE=Francois;9918]Brainstorm does not do any volume registration at this time.
If you are trying to import two volumes for the same subject (or one structural MRI and one atlas), the two volumes need to be coregistered first (same dimensions, same orientation). You need to do this in an external program.
We are planning to add tools for manipulating MRI volumes more easily, but we currently do not have the necessary resources to develop them.
http://neuroimage.usc.edu/brainstorm/Next#Anatomy[/QUOTE]
Aha thank you!
So I should do the coregistration in freesurfer (as example), than I enter the resulted volume in the anatomy folder?
Yes. But I’m sorry I cannot guide you more precisely, I don’t know how to do it myself yet.
[QUOTE=Francois;9920]Yes. But I’m sorry I cannot guide you more precisely, I don’t know how to do it myself yet.[/QUOTE]
Thank you very much !! 
Thank you Francois, it helped me
Dear Francois,
After doing the coregistration between the anatomical surface and the .nii atlas ROIs, the loading of the atlas using (volume mask or mri files) doesn’t generate any error.
However, the resulting atlas is in the picture 1, while I should obtain as the picture 2 …
Do you have any idea?
Hello,
Can you send me these files?
(upload them somewhere, eg dropbox, then send me the link in a separate email - click on my username on this forum)
Please explain what each file means and what menus you use to load them.
Thanks
Francois
Hi Aya,
The .nii file you sent me is not correctly labelled: the volume was smoothed or compressed, maybe because of this coregistration you mention.
One region is not represented by one value only, as it should be, but by tens of values to make the edges look smooth in a viewer.
The procedure you used to generate this atlas is not correct, you cannot use this file as an atlas in any program (not only Brainstorm).
Please note that Brainstorm does not offer any proper way of importing volume atlases as surface ROIs. The function that you were trying to use is a very rough approximation that attributes a vertex of the surface to a region if it is contained inside the region. It sounds OK, but creating actual surface atlases where all the vertices are attached to a region is a lot more complicated than this. So even if you get a correctly registered file, you cannot expect good results in Brainstorm.
You should maybe try to do this step outside of Brainstorm (in FreeSurfer?) and then import the surface atlases in Brainstorm. You could try asking these questions to a team specialized in volume processing and atlases.
Sorry for not being able to help you further.
Francois