It's possible, but this isn't what preborder.sh does. First, its depth maps
are generated from a version of the midthickness that has been stereotaxically
registered to WashU's 711-2B space (because this is what the probabilistic
volumes representing the sulci are in). But also it has already been resampled
to 73730, for complex reasons.
Still, preborder.sh's generate_depth subroutine could be applied to the orig
space, native mesh surface:
> generate_depth() {
> CASE=$1
> HEMISPHERE=$2
> HEM_FLAG=`echo $HEMISPHERE | cut -c1 | tr '[:lower:]' '[:upper:]'`
> SPEC=Human.$CASE.$HEM_FLAG.73730.spec
> echo Generating depth for $CASE $HEMISPHERE
> # need to create volume to generate segmentation from fiducial
> caret_command -volume-create 176 208 176
> Human.$CASE.$HEM_FLAG.segmentation.nii
> caret_command -volume-set-origin Human.$CASE.$HEM_FLAG.segmentation.nii
> Human.$CASE.$HEM_FLAG.segmentation.nii -88.0 -123.0 -75.0
> caret_command -volume-set-spacing Human.$CASE.$HEM_FLAG.segmentation.nii
> Human.$CASE.$HEM_FLAG.segmentation.nii 1.0 1.0 1.0
>
> # generate segmentation from fiducial
> FIDUCIAL=deformed_Human."$CASE.$HEM_FLAG".Midthickness_711-2B.coord
> TOPO=Human.$CASE.$HEM_FLAG.CLOSED.73730.topo
> caret_command -surface-to-segmentation-volume $FIDUCIAL $TOPO
> Human.$CASE.$HEM_FLAG.segmentation.nii
> # generate hull volume, surface, depth (shape), and geography paint files
> caret_command -surface-identify-sulci $SPEC $HEMISPHERE
> Human.$CASE.$HEM_FLAG.segmentation.nii $TOPO $FIDUCIAL $FIDUCIAL NIFTI_GZIP
> }
Note that Caret's depth is computed on the midthickness. If you used something
like rh.white or rh.pial, instead of generating a volume from the midthickness,
you'd get a qualitatively different map.
Currently, preborder.sh doesn't support deforming back from atlas to native
mesh, which is a two-step process (downsampling, then registration). The
registration step is two-way, but not the downsampling step. A few more steps
could be added to make it that way, but the lab is moving to fs_LR, and it
sounds like you want your depth maps sans spatial normalization.
On Feb 17, 2012, at 2:11 AM, j janssen wrote:
> Hi Donna,
>
> context: advances in streamlining FreeSurfer and Caret processing
> (PALS_B12.LR.zip).
>
> goal: I have my subject's T1 ran through FreeSurfer's recon-all. for
> this same subject i want to use Caret to obtain native space sulcal
> depth maps. thereafter i want to take this Caret native space sulcal
> depth map back to FreeSurfer land such that it has an identical format
> as FreeSurfer's ?h.thickness maps. so, in FreeSurfer land i would have
> a thickness measurement for each vertex (?h.thickness) and a sulcal
> depth measurement for that same vertex (Caret->FreeSurfer sulcal depth
> map).
>
> question: is this possible?
>
> thanks!
> -joost
> _______________________________________________
> caret-users mailing list
> [email protected]
> http://brainvis.wustl.edu/mailman/listinfo/caret-users
_______________________________________________
caret-users mailing list
[email protected]
http://brainvis.wustl.edu/mailman/listinfo/caret-users
