Hey there,
I was excited to find out that Freesurfer has two atlases built in which do cortical parcellation as I am trying to do analysis on the precentral gyrus ROI only. However, when creating a mask showing the pre-central gyrus parcellation, I noticed that in the middle of the gyrus the algorithm fails and skips to the next gyrus over (the post-central gyrus). I added an image showing the failed parecellation (not sure if it shows up).
Is there any solution around this? Is there where I can change parcellation options to maybe help better restrict the automatic algorithm? Or is there an easy way to redirect/train the algorithm?
Thank you very much for any advice and your time,
Omer Tal, M.S.
UCSD Center for Functional MRI
Hi Omer,
there is actually a work-around for this kind of thing. You need to create a central sulcus label in tksurfer:
1. Load the inflated surface with sulc as the "curvature" overlay. 2. Draw a path across the central sulcus near the vertex to prevent the filling (see below) from escaping the sulcus. 3. Click on a point in the central sulcus. 4. Click the "custom fill" button, and do a fill with "last vertex clicked", "up to and including paths" and "up to and including different curvature". 5. Save the resulting label as lh.central_sulcus.label or rh.central_sulcus.label 6. Rerun mris_register using:
set sd=$SUBJECTS_DIR/$subject set ldir=$SUBJECTS_DIR/$subject/label set sdir=$SUBJECTS_DIR/$subject/surf mris_register -l $ldir/lh.central_sulcus $FREESURFER_HOME/lh.destrieux.simple.2009-07-29.gcs S_central...
where ... means the rest of the normal arguments.
cheers Bruce
On Thu, 15 Sep 2011, Omer Tal wrote:
Hey there,
I was excited to find out that Freesurfer has two atlases built in which do cortical parcellation as I am trying to do analysis on the precentral gyrus ROI only. However, when creating a mask showing the pre-central gyrus parcellation, I noticed that in the middle of the gyrus the algorithm fails and skips to the next gyrus over (the post-central gyrus). I added an image showing the failed parecellation (not sure if it shows up).
Is there any solution around this? Is there where I can change parcellation options to maybe help better restrict the automatic algorithm? Or is there an easy way to redirect/train the algorithm?
Thank you very much for any advice and your time,
Omer Tal, M.S.
UCSD Center for Functional MRI
p.s. sorry, then you'll need to rerun the cortical parcellation steps afterwards (mris_ca_label). The full mris_register command line would be:
set sd=$SUBJECTS_DIR/$subject set ldir=$SUBJECTS_DIR/$subject/label set sdir=$SUBJECTS_DIR/$subject/surf mris_register -l $ldir/lh.central_sulcus \ $FREESURFER_HOME/lh.destrieux.simple.2009-07-29.gcs S_central \ $sdir/lh.sphere $FREESURFER_HOME/average/lh.average.curvature.filled.buckner40.tif $sdir/lh.sphere.reg
On Thu, 15 Sep 2011, Omer Tal wrote:
Hey there,
I was excited to find out that Freesurfer has two atlases built in which do cortical parcellation as I am trying to do analysis on the precentral gyrus ROI only. However, when creating a mask showing the pre-central gyrus parcellation, I noticed that in the middle of the gyrus the algorithm fails and skips to the next gyrus over (the post-central gyrus). I added an image showing the failed parecellation (not sure if it shows up).
Is there any solution around this? Is there where I can change parcellation options to maybe help better restrict the automatic algorithm? Or is there an easy way to redirect/train the algorithm?
Thank you very much for any advice and your time,
Omer Tal, M.S.
UCSD Center for Functional MRI
freesurfer@nmr.mgh.harvard.edu
-
Bruce Fischl -
Omer Tal