display Cortical Surface Parameterization on gyrus mesh

Questions about BrainVisa usage and installation

Moderators: denghien, riviere

Post Reply
kl1597
Posts: 25
Joined: Thu Oct 06, 2005 2:21 pm

display Cortical Surface Parameterization on gyrus mesh

Post by kl1597 »

Dear Brainvisa expert and Dr. Olivier Coulon,
I use Cortical Surface Parameterization tool and have some questions,
1. could I apply the final Parcellation result on gyri mesh ? it look like on the white matter surface but I hope I can display it on gyrus surface
2. what's the mapping area name ? it there any atlas on pdf document ?
3. Is it possible to calculate each area's volume ?
thanks your replay
kl chen
User avatar
Olivier Coulon
Posts: 176
Joined: Fri Feb 27, 2004 11:48 am
Location: MeCA research group, Institut de Neurosciences de La Timone, Marseille, France
Contact:

Re: display Cortical Surface Parameterization on gyrus mesh

Post by Olivier Coulon »

Hello,
the process that provides the cortical surface parameterization is on user level 2 (expert) in Brainvisa, which means that it is not really supported. Not many people use it yet and it is still in development. Nevertheless :
1. could I apply the final Parcellation result on gyri mesh ? it look like on the white matter surface but I hope I can display it on gyrus surface
2. what's the mapping area name ? it there any atlas on pdf document ?
3. Is it possible to calculate each area's volume ?
1 - not really. I guess you mean the grey matter mesh. Everythng is done on the white matter surface there is not (yet) an easy way to propagate to the grey matter mesh.
2 - no name for the parcellation sets (yet, again). But it will come. The only good thing is that each parcel has a number (represented by the color you see on the screen) and the number is always the same across subjects.
3 - no (not yet). You can compute the surface pretty easily but it involves a little bit of programming, by getting all the triangles that belong to a particular area and summing up their surface area...

All this will be part of the process in a more or less close future but you will have to be a bit patient.

Olivier
Olivier Coulon
Institut de Neurosciences de La Timone,
Aix-Marseille Université,
Marseille, france
https://meca-brain.org
kathleenACE
Posts: 9
Joined: Fri Apr 17, 2009 10:28 pm

Re: display Cortical Surface Parameterization on gyrus mesh

Post by kathleenACE »

You can project the result of the white matter parcellation onto the hemimesh with a fusion:
1. make the fusion of the .tex and white matter mesh as you would to view it
2. select that fusion and the hemimesh and make another fusion, but change the type of fusion to interpoler
3. view the interpoler mesh in a 3D window (this can take a little extra time to load, so be patient)
4. to save this as a .tex, right-click on it in anatomist and select file-->export texture (this way you can open it again)

I am still working on getting ROIs from this for gathering measures... if anyone figures that out please post it.
If it can be converted to a volume (.ima) file, then getting ROIs would be easier.
I am looking for a command to convert the .tex to a volume, or to a graph and then to a volume, but none of the commands in the bin have worked for me yet.

Hope this helps,
Kathleen
User avatar
riviere
Site Admin
Posts: 1361
Joined: Tue Jan 06, 2004 12:21 pm
Location: CEA NeuroSpin, Saint Aubin, France
Contact:

Re: display Cortical Surface Parameterization on gyrus mesh

Post by riviere »

Hi,

Yes, you're right, the interpoler may help projecting white mesh data onto a grey mesh. But it is a quite basic tool, it just interpolates one mesh to another one "not too far", by projecting each vertex of the first mesh to the nearest triangle of the second one, and performs a linear interpolation to compute values. So don't expect very accurate results, especially on high curvature regions. Moreover it is largely sub-optimal on calculation time (as you have noticed). It was writen 10 years ago to be a little tool to project EEG data (a mesh with, say, 128 vertices) on a head mesh, so did not need very clever optimization at that time...

Now to do the contrary: print a mesh texture into a volume, I am afraid we don't have tools designed for it: it would need a kind of rasterization algorithm to "draw" (interpolate) triangles in a voxels grid. This kind of algorithm is very classical (every graphic card use this, even for 2D line drawing), but we just never needed one before and thus don't have one in our tools, as we are primarily working on volumes (voxels), then mesh objects we extract, then later some surfacic processings were developed, but we ususally don't convert surfacic objects back to volumes.
There might exist software which can perform this I guess, because rasterization algorithms are described in every computer graphics book...
Otherwise you may try some dirty crappy simple things based on our tools, if you program a little bit (in python especially, it would be done in 3-4 lines): "print" each vertex mesh in the voxel it is located in, then perform some kind of closing on the volume to fill holes (assuming triangles are not too large) - but this is a "quick and dirty", not to say ugly, solution...

Denis
Post Reply