Hello BST team,
I would like to use MarsAtlas as ROI.
I downloaded a nii file this link.
http://meca-brain.org/software/marsatlas-colin27/
Next
In Scout Tab, Atlas > Load atlas... I selected colin27_MNI_MarsAtlas.nii. (file type: Volume mask or atlas(MNI space))
The result is this figure.
I think this atlas is unclearly.
What should I do to solve the problem?
Thank you for your help.
mitsuo
Hello,
It is not a good idea to try to load a volume atlas as a set of surface ROIs. The result is terribly imprecise, as you can see.
The MarsAtlas parcellations are now loaded when importing a BrainVISA anatomy folder:
https://github.com/brainstorm-tools/brainstorm3/commit/979771fc0ff98ffe45a2198e04f50ba7373fb1f0
You could try two things:
If you can find a full segmentation of an MNI template generated with a recent version of BrainVISA: load it with the menu "Import anatomy folder". https://neuroimage.usc.edu/brainstorm/Tutorials/ImportAnatomy#Import_the_anatomy
Process the ICBM152 volume with BrainVISA yourself and then import it:
http://www.bic.mni.mcgill.ca/ServicesAtlases/ICBM152NLin2009
https://neuroimage.usc.edu/brainstorm/Tutorials/SegBrainVisa
Hello,
I did the importation of the MarsAtlas on the Colin27 template.
The BST template is available here.
Note that the Colin27 there is slightly different from the one included in Brainstorm (tailored for NIRS).
Best.
Thomas.
@tvincent Thanks!
Any reason why you are still using the Colin27 brain instead of the MNI 152? (2009)
Thanks for reply everyone.
I have only MEG data(don't have fMRI data).
Can I do these process?
Thank you for your reply.
mitsuo
There is no fMRI involved anywhere here.
You should be able to download the .zip file that Thomas sent.
Create a new protocol, copy all the files in the package to the anat/@default_subject folder.
Then in Brainstorm, right-click on (Default anatomy) > Reload.
After that you should be able to use this as an anatomy template for estimating sources for your MEG recordings. If you have access to the head shape of the subjects, you would improve the source localization by warping the template to the head shape of the individual subjects.
https://neuroimage.usc.edu/brainstorm/Tutorials/TutWarping
It's because we needed a segmentation of the extra-cortical layers that only the Colin27 template was providing (among the "common" templates). It is used to evaluate light propagation from the scalp.
Thanks for your advice.
I have done the process you taught me.
This figure is clearer than before.
However, some scouts contain other scout's vertices.
What should I do to solve this problem?
Thank you for your help.
mitsuo
From the Brainstorm templates?...
Where does this segmentation come from?
Thank you for reply.
Sorry, I don't know where does this segmentation come from.
How can I know where does this segmentation come from?
mitsuo
I meant: what files did you have in input and what menus did you use to import them and produce the figure you posted?
Thank you for reply.
First, I did warping the anatomy templates for subjects (templates: Colin27 4NIRS).
https://neuroimage.usc.edu/brainstorm/Tutorials/TutWarping
Then, I did less vertices... on white from 25003 V to 7500 V.
Next, I did compute head model and sources[2018].
After, I applied the MarsAtlas.
(Previous figure was chose the cortex_7500V so the figure is little different before one.)
Did I answer your question?
Sincerely,
mitsuo
@mitsuo What do you mean with "After, I applied the MarsAtlas"?
@tvincent Did you produce these atlases? how?
I meant I chose the MarsAtlas at Scout Tab.
Hello,
I used a nearest neighbor approach to interpolate the atlas labels volume onto the cortical mesh.
I took the labels volume from the MarsAtlas website (https://meca-brain.org/software/marsatlas-colin27/). I coregistered it on the Colin27 MRI that is part of the Colin27_4NIRS template, using the Colin27 MRI provided in the MarsAtlas website.
To be precise, there is no scout that contains vertices from another scout. But I agree that some vertices could be better assigned. I will force all regions to be connected so that there is no "isolated" vertices.
I should be able to do that soon.
A much better approach would be to use directly the MarsAtlas surface maps, now they are loaded automatically by Brainstorm when using the menu "Import anatomy folder" to load a BrainVISA segmentation.
This would give you all the benefits of the surface-based registration, without any additional interpolation needed.
