Dear Martin,

As far as I know those meshes were prepared using software called
BrainVISA based on a brain normalised to the template, but not on the
template brain. This might explain your findings. We are actually
about to replace them with new set of meshes that were derived from
the MNI template. I can send you those meshes off the list and it'll
be quite useful if you re-check that they correspond to the template
MRI.

Best,

Vladimir



On Tue, Feb 3, 2009 at 9:59 PM, Martin Luessi <[log in to unmask]> wrote:
> Hi,
>
> I was wondering how the template cortical meshes that are included in
> SPM8b (they are the same as in SPM5) were obtained. The meshes are
> used for EEG 3D source reconstruction and are stored in the
> "EEGtemplates" sub directory. There are four mesh sizes available,
> with 3004, 4004, 5004, and 7204 vertices, the corresponding files are
> named "wmeshTemplate_3004d.mat" etc.
>
> I am trying to use the template meshes for EEG source reconstruction
> and fMRI/EEG fusion. For this application it is important that the
> vertices are located on the cortical surface (gray matter surface). I
> have been comparing the meshes with high resolution structural MR
> image that is included with the EEGtemplates ("HighRes_smri.nii" in
> SPM8b) and it seems like the mesh does not correspond well to the
> structural MR image. More specifically, the mesh seems to be "too
> small", i.e. many vertices are within the white matter of the sMR
> image.
>
> It would be very helpful for me to know how these meshes were
> obtained, i.e. which structural MR images and what algorithms were
> used.
>
> Thanks,
>
> Martin
>