Hi Arselle
the node_spacing is the mm resolution of the atlas Markov priors and class-conditional densities, the prior_spacing is the same for the classwise spatial priors
the equivalent for the surface is mris_ca_train. It takes annot files as input
cheers Bruce
On Mon, 4 Jun 2018, arsene ella wrote:
External Email - Use Caution
Dear All,
I have a set of 7 NIIFTI images (0.5 mm isotropic) :
- 2 average brain images (one T1W_MPRAGE and one T2W_SPACE template )
- 3 Tissue Probability Maps (Gray Matter, White Matter, CSF)
- 2 Atlases (One for Gray matter structures and another for Internal brain structures (or subcortical structures) )
segmented from mean images, labelled and annotated.
My question is: How to used these images to build Freesurfer-Like Atlases (.gca and .gcs) ?
For the .GCA, I don't really know what to set as Flagged Arguments...
mri_ca_train -node_spacing 4 \ # ? -prior_spacing 2 \ # ? -parc_dir esegatlast \ # Is that the directory of my segmented subcortical structures? -xform talairach_one.m3d \ # All the files are aligned, so which transform shold I set here? -input flash30_avg.mgz \ # Should I set the T1W-MPRAGE average here? -input flash5_avg.mgz \ # Should I set the T2W-SPACE average here? $SUBJECTS multi_one.gca
What about the .GCS?
Thanks in advance,
Arsene