It is better to use mri_label2vol using the registration matrix for the functional. You can set the "fill threshold" to control the percent of a segmentation that must be present in a voxel to declare that voxel to be in that segment. I would not set this to 1 (ie, 100%) as this can lead to some artifacts since it just counts the number of anat voxels that land in the functional voxel. Set it to something like .85 or .9.
doug
Andrew Ward wrote:
Hi Freesurfers,
I have created a binary mask of the ventricles using mri_binarize using data from aparc+aseg.mgz. I want to resample this mask into native functional space, but I want to include only functional voxels that are 100% covered by by the structural mask. Because the anatomical mask is written with voxels that are 1x1x1 and our functional data is re-sliced at 3x3x3, there are functional voxels that are more than 50% covered by the anatomical mask that I want to exclude. I used the command:
mri_vol2vol --mov brainmask.mgz --targ ventmask.nii --ing --reg register.dat --interp nearest --o func.ventmask.nii
This properly resampled the mask into functional space, and maintained its binary character, but it looks to be a bit too permissive -- although I could just be seeing things.
Any ideas?
Thanks for your help,
Andrew Ward Brigham and Women's Hospital (BBNG) Massachusetts General Hospital (GRU) Athinoula A. Martinos Center for Biomedical Imaging 149 13th St., Rm 2689C Charlestown, MA 02129 Phone: 617-726-6211 Fax: 617-726-5760
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer