Hello everyone. I'm trying to find a file of the cortical parcellation MarsAtlas (auzias 2016), because I would like to use it in a whole brain connectivity analysis. Does anyone have an .annot file or other format so that I can load it directly on a surface in brainstorm?
thanks
In order to get the MarsAtlas parcellation, you should process your subject's MRI with BrainVISA, or a mixed pipeline FreeSurfer-BrainVISA:
https://neuroimage.usc.edu/brainstorm/Tutorials/SegBrainVisa
http://meca-brain.org/software/hip-hop/
Unless @ocoulon @guillaume_auzias or @Brovelli have other suggestions?
Indeed I have already tried to segment in BrainVisa the T1 mri ICBM152 (beacuse I use the default anatomy), unfotunately I did not find the MarsAtlas. The segmentation worked correctly with brainvisa, although when I imported the surfacese in brainstorm, in the brainvisa folder I did not find the atlas to import in brainstorm but only left and right hemisphere surface file. I downloaded a smoothed surface with the atlas from here
https://www.dropbox.com/s/1l35wxu09g44un5/hiphop138-Template.tgz?dl=0
It works fine, but the only problem in this case is that if I want to use individual anatomy and to overlay the atlas I don't know how to do it. The segmentation in brainvisa did not produce automatically the marsatlas in the way it should do by following brainstorm tutorial.
Do you have any suggestion about how to solve the segmentation issue of brainvisa?
I could try to generate the surface from the segmented colin27_MNI normalized MRI file availabe here:
https://www.dropbox.com/s/ndz8qtqblkciole/MarsAtlas-MNI-Colin27.zip?dl=0
I shiuld have the mars atlas inside, and maybe generating the surface from it could solve my problem.
Do you think it could work?
Thanks
Indeed, the Brainstorm tutorial for BrainVISA is wrong, it does not compute the MarsAtlas parcellation.
In the Morphologist options, you need to select the checkbox Sulci graph morphometry by subject.
And then you need to run a separate process in BrainVISA: Cortical surface > Anatomy > Hip-Hop Cortical Parameterization
I will update the tutorial soon.
In the meantime, let us know if it works.
Thanks for reporting this issue
Hi Francois, a small update on the marsatlas. I followed your suggestion. To obtain the Atlas in the morphologist option you have to select not only the Sulci graph morphometry by subject but also Sulci recognition and Sulci recognition with SPAM. After that you can go through the Hip-Hop parametrization. Strangely I obtained the atlas only for the White files which I supposed to be the white matter, maybe I did something wrong, but I thought that everything was correct since I obtained the mesh for gray and white matter.
Hope this could be helpful in some way.
best
Gian Marco
To obtain the Atlas in the morphologist option you have to select not only the Sulci graph morphometry by subject but also Sulci recognition and Sulci recognition with SPAM .
Indeed, I forgot to mention it.
Do the instruction I added in this tutorial look good to you?
https://neuroimage.usc.edu/brainstorm/Tutorials/SegBrainVisa#Running_Morphologist
Strangely I obtained the atlas only for the White files which I supposed to be the white matter, maybe > I did something wrong, but I thought that everything was correct since I obtained the mesh for gray and white matter.
Yes, this is what this method does, it is the expected output. It was already documented:
https://neuroimage.usc.edu/brainstorm/Tutorials/SegBrainVisa#Cortical_parcellations
You should then compute your sources on the white surface: this is not the default in Brainstorm, but some other research groups prefer this grey-white interface for source estimation.
Projecting the atlas from the white to the pial surface using Brainstorm would not be accurate, as we do not have access to a parametrization of the pial surface (unlike with BrainSuite, FreeSurfer or CAT).
Yes I think that now the instruction is clearer than before.
I didn't understand completely before your explanation, thanks!