I have extracted head.mesh, Lhemi.mesh and Rhemi.mesh from BrainVisa and I have imported them together with MRI as the individual anatomy of one subject in Brainstorm. I followed the tutorial steps. I checked the registration of the meshes with the MRI (figure attched) and it is fairly good indeed. The problem is that I want to include also the cerebellum in my analysis as I expect some sources to be there.
Is there a way to have a corresponding .mesh file also for the cerebellum from BrainVisa? Does anybody know it?
I’m sorry I don’t know the answer to this question. If you can get a mesh for the cerebellum, you’d be able to use it for source reconstruction in Brainstorm. But how do you get this?..
You should ask the authors of BrainVISA, or you can have a look at other programs that are doing a similar job (FreeSurfer, BrainSuite). I guess that at least one of them is going to give you a tesselated brain mask that includes the cerebellum, or a separate mesh file for the it.
Not totally positive, but it’s from 2008. They may have an update as of today. Otherwise, ask the same question to the Freesurfer and Brainsuite community. Please keep us posted as indeed, we’d like to include the cerebellum to the brain templates featured in Brainstorm.
However, I found something by myself. Maybe it also useful to you. I used the Ana Get Opened Whole Brain Surface tool in Brainvisa (components_obsolete -> segmentation) and I extracted a brain_mesh where the cerebellum is included. Then in Brainstorm I imported all Lhemi.mesh, Rhemi.mesh and brain.mesh and I merged the 3 meshes. I took a fairly good alignment with MRI at the end. I also tried only the brain.mesh alone but the alignment was much worse.
I know it is not the proper way, but this is what I could think.
[QUOTE=Sylvain;3485]Sounds like a workable solution; thanks for sharing Chrysa!
How did you merge the 3 meshes in the end?[/QUOTE]
I imported as surfaces the 3 meshes, I downsampled the number of vertices as in the tutorial, and I selected the three of them and rightclick–> merge surfaces. The new surface includes also the cerebellum and I tested the alignment with MRI as seen in the figure I attached.
From the Brainvisa forum they proposed to use some lower level functions to get a cerebellum mesh with high resolution. Here is the thread: http://brainvisa.info/forum/viewtopic.php?f=2&t=1613
But as it is written, it is not so easy… So, I ll go on with the ‘workable’ solution and keep you updated.
I am just wondering how the merged surfaces would look in the end. I am afraid that merging the cortical surfaces this way would yield weird-looking sulci and gyri if the two versions of the tessellations (W and w/o cerebellum) would not perfectly align. Maybe you could send out a screenshot of the 3D rendering of the resulting surface, with the ‘Sulci’ option activated in the Surface panel, to mark brain convolutions better. (right click over the cortical surface of your Subject in the database and select ‘Display’)
Indeed, I’m not sure it is a good idea to go merge the tesselated brain mask + the two hemispheres. You would end up having two overlapping layers for both hemispheres.
If you want to use the tesselated brain mask only (brain.mesh), it’s ok, but in this case you should reconstruct the sources without orientation constrained (when you compute the inverse model, select wMNE, then Expert mode, Ok, then select the option “Unconstrained”)
If not, they were offering an interesting alternative: having a mask/mesh of both the cerebellum and the brainstem. You could go this way, it’s probably not a problem if you have other structures in your mesh…
I found out that Freesurfer can produce a cerebellum surface with the command mri_tessellate. I thus believe that you can get all left/right hemispheres and cerebellum using this software. Do you want me to find out more details and let you know?
Sylvain, I attach the sulci image I get. Is it so bad? Would you use it in the next steps?
Francois, the brain.mesh does not provide good alignment with the MRI at the end, so i prefer not to use only this mesh. Is it BrainVisa that offers the mask with both cerebellum and brainstem? I could not find it.
Dimitrios, thanks for the info. Yes, if you find more details, i would like to try FreeSurfer also. It can run only on Linux, right? This is a problem but I can set up a virtual pc. Will it be easy then just to import the meshes in Brainstorm in a windows running PC? Say hi to Christos also!
Sylvain, I attach the sulci image I get. Is it so bad? Would you use it in the next steps?
You should definitely not use this surface, it has two layers on top of each other for each hemisphere.
Francois, the brain.mesh does not provide good alignment with the MRI at the end, so i prefer not to use only this mesh.
You said that the surface you created and that you show on your screenshots is a combination of lhemi.mesh, rhemi.mesh and brain.mesh.
Try using just brain.mesh...
Is it BrainVisa that offers the mask with both cerebellum and brainstem? I could not find it.
On the BrainVISA forum, Denis Riviere replied: "Getting a segmentation of the cerebellum + brainstem is easy."
You can probably ask him more details about that
I would like to try FreeSurfer also. It can run only on Linux, right?
From the FreeSurfer website, you can download a complete virtual machine that includes a full install of FreeSurfer. You can download this on a PC or a MAC are run it using VirtualBox.
With this, you can run your FreeSurfer MRI segmentations from your windows PC, then use the surfaces on your Windows Brainstorm install.