7 Apr
2011
7 Apr
'11
10:18 a.m.
More about Beamforming files
In a previous mail-list I was told that the *spheres.txt and
*transform.txt file are obtained after coregistration
of the MEG data to the MRI structural before beamforming. So I imagine
that transform.txt matrix takes the NAI voxel values to
the MNI space with a 5mm or 2mm volume grid, from the individual/subject
brain to the MNI standard. Am I right? Is it MNI standard or the MNI
space of the individual structural MRI?
But anyway, (and this is a supposition) coregistration should be done
first between MEG and the individual MRI structure to perform Beamforming
in world/real or physical coordinates (right?). So there should be a
previous space transform in world coordinates to perform beamforming,
then the NAI values are obtained in world coordinates and finally
transformed to MNI space together with the individual structural MRI.
Am I right about this?
The reason why I am to eager to know this, is because I am doing
Beamforming for my project using python-vtk, and so far I have learned
how to extract the cortical voxels of individual structrual MRI using
BET-FAST from FSL. But in order to run my beamforming I need to
transform the structural voxels (in MNI) to world coordinates and also
do coregistration between the fiducial skin points, which I imagine were
obtained using the stylus pen for 3D modelling during the MEG
acquisition with the surface of the subject skin from the MRI.
I was wondering if the transform.txt file does the coregistration
skin-points-to-skin-surface for the Beamforming, Now I know that it does
not.
Maybe I will have to program my own coregistration algorithm anyway. I
only wanted to know If there were a less painful way to do it.
My fifth question is the following. According to the ynic-wiki, it says
that I am able to define any voxel in MNI coordinates to define virtual
electrodes using the Beamforming functions. The problem is that the wiki
does not say how to define this coordinates. From where can I take them?
From the structural MRI or from the MRI standard? Can
they be only the
cortex or any voxel inside the brain?
And my last question is about the sphere.txt file. It is composed of 248
lines, they should be the MEG coils, but I can not imagine the meaning
of the rest of the four columns. I was thinking they where x,y,z
coordinates, but there is a fourth one that tells me I am wrong.
Many thanks,
Luis R. Peraza