External Email - Use Caution
Hello,
I have tried multiple strategies to create labels in subject space from a couple subcortical (brainstem and cerebellum) volume-based ROIs (masks) defined from a thresholded statistical map. The ROI file is a simple binary mask in MNI space. However, the results are misaligned and appear as a grid of dots.
First I tried using mri_vol2vol to go from MNI space to FS space (I tried using both the fsaverage brain and the mni205 brain). I previously ran the MNI152 brain through recon_all, per advice I found in the archives. For example, using the below command:
mri_vol2vol --mov ${ROI_DIR}${roi}.nii.gz --s MNI152_T1_2mm \ --targ $FREESURFER_HOME/average/mni305.cor.mgz --m3z talairach.m3z \ --o ${ROI_DIR}${roi}_FSvol.nii.gz --nearest
Since the MNI space ROIs also overlaid in FS space, I also tried skipping this first step and going straight to the next step, which was to use mri_vol2label to create the label file. For example, like this:
mri_vol2label --i ${ROI_DIR}${roi}_FSvol.nii.gz --id 1 --l ${ROI_DIR}${roi}_FSvol.label
Then I used mri_label2label to move the label to each subject's space, like so: mri_label2label --srcsubject fsaverage \ --srclabel ${ROI_DIR}${roi}.label --trgsubject ${subfolder} \ --trglabel ${SUBJECTS_DIR}/rois/${sub}_${roi}_FSvol.label --regmethod volume
I also tried moving the label output of this final step (in subject space) back to a volume, though this doesn't seem necessary, as I could extract the data from the label file:
mri_label2vol --label ${SUBJECTS_DIR}/rois/${sub}_${roi}_mni.label \ --o ${SUBJECTS_DIR}/rois/${sub}_${roi}_FSvol.mgh --identity --temp ${SUBJECTS_DIR}/${subfolder}/mri/T1.mgz
I have attached a screenshot of the label file for the cerebellar ROI (in yellow) in subject space, overlaid on the subject's T1 from: /subject_directory/mri/T1.mgz (I apologize that it's not the best view, as I'm working from home and it's very slow using the GUI to change the view). I also apologize in advance if I've left out critical info to answer this question - please let me know.
I appreciate any advice you might have! This seems like it should be a pretty common, straightforward process, yet I have not been able to find anything that works.
Thank you,
Tammi RA Kral, PhD
Center for Healthy Minds University of Wisconsin-Madison she/her
