Hi FreeSurfer experts,

I have two different ways to do functional connectivity, and was wondering which one is optimal since they produce different results. The first is handed down to me from a previous coordinator. However a few of the steps seemed overly complicated, so I looked into a more direct method. Both methods use single vertex as a seed, so we smooth in order to include enough vertexes for fcseed-sess.

Option 1 (the old version):

Creating the mask, smoothing, and binarizing it: ~ mri_label2label --s fsaverage --regmethod surface --hemi lh --srclabel $SUBJECTS_DIR/$subject/label/$label.label --trglabel $SUBJECTS_DIR/$subject/label/junk.label--outmask $SUBJECTS_DIR/$subject/mri/{$label}_sm00.mgh ~ mris_fwhm --i $SUBJECTS_DIR/$subject/mri/{$label}_sm00.mgh --fwhm 4 --smooth-only --o $SUBJECTS_DIR/$subject/mri/{$label}_sm04.mgh --s fsaverage --hemi lh ~ mri_binarize --i $SUBJECTS_DIR/$subject/mri/${label}_sm04.mgh --min 10e-10 --o $SUBJECTS_DIR/$subject/mri/{$label}_sm04.10e-10.bin.mgh

Switching to dev in order to convert the mask into a volume: ~ source /usr/local/freesurfer/nmr-dev-env ~ mri_surf2vol --so $SUBJECTS_DIR/fsaverage1/surf/lh.pial $SUBJECTS_DIR/$subject/mri/${label}_sm04.10e-10.bin.mgh --subject fsaverage --o $SUBJECTS_DIR/$subject/mri/{$label}_sm04.10e-10.bin.vol.mgz

We then move back into 5.3 and use this volume as segmentation in fcseed-config and -sess

Option 2:

Turning label into a volume, smoothing it, and binarizing it: ~ mri_label2vol --label $SUBJECTS_DIR/fsaverage1/label/{$label} --temp $SUBJECTS_DIR/fsaverage1/orig.mgz --regheader $SUBJECTS_DIR/fsaverage1/orig.mgz --hemi lh --subject fsaverage --proj frac 0 1 .1 --o $SUBJECTS_DIR/{$subject}/{$label}.vol.mgz ~ mri_fwhm --i $SUBJECTS_DIR/{$subject}/{$label}.vol.mgz --fwhm 4 --smooth-only --o $SUBJECTS_DIR/{$subject}/{$label}.sm4.vol.mgz ~ mri_binarize --i $SUBJECTS_DIR/{$subject}/{$label}.sm4.vol.mgz --min 10e-10 --o $SUBJECTS_DIR/{$subject}/{$label}.sm4.bin.vol.mgz

We then use this volume as segmentation in fcseed-config and -sess.

I would prefer to use the second method since it is more straightforward, but I wanted to make sure that a) it is sound and b) it is more optimal than option 1.

Thank you for your help!

Anais Rodriguez-Thompson Clinical Research Coordinator Brain Genomics Laboratory Martinos Center for Biomedical Imaging Massachusetts General Hospital Phone: (617) 643-3215