External Email - Use Caution
Dear FreeSurfer Team,
Hi. I am writing this email to ask if there is any way to do functional ROI analysis on the both hemispheres at the same time, but not separately. My research question is to investigate visual hierarchy system under certain fMRI tasks using surface-based analysis. I am not interested in differences between hemispheres. There was the same question previously (https://secure-web.cisco.com/1JFM2A4prS9i4P3MGZiI1ScaCdFi7AFRFt2tlUzGfYgNkr5...https://secure-web.cisco.com/1n5iUWdnTAoX8uAScZjw9vgkzDWKHItczK2oXOxDKgvBSjo10mym4SxVhJdbv04lsbVGAPlmlTHKnW0fyN3HhpsyFvGmk1ttO4Hp9mfdogB-xN_r77UkK8idhutNCy4FFf9S2Vi8sBBs2GMoH8BICwOfm0IkBMEIrUD6VFSoXonnxlerlwDqVPlO07e2UnO-S31iuW1SiJsCF6vyfHXZFuybOMI_5oUVmz0yExNXQLgjtBjC25nYidKVYgTlpse0FqTMOad3hyadg8vhVUi9JnBdU5P8LXgoyG30OCyovzwRYGIELos6mjGCZBTQmACni_0ePaR-hv3UzNbGZS_5Z5Q/https%3A%2F%2Fwww.mail-archive.com%2Ffreesurfer%40nmr.mgh.harvard.edu%2Fmsg18681.html) but the answer didn't help. What I understood from the post was that I could do ROI analysis on both hemispheres just like I did it on one hemisphere. So, I tried functional ROI analysis using labels from the both hemisphere (e.g., whole brain V1, not left hemisphere V1). I hypothesized that the results of ROI analyses with left hemisphere and right hemisphere would be the same If it really works. On the other hand, it produced errors and I even couldn't compare the results.
Here is what I tried:
# merge lh and rh labels mri_mergelabels -i lh.wang2015atlas.V1d.label -i lh.wang2015atlas.V1v.label -i lh.wang2015atlas.V2d.label -i lh.wang2015atlas.V2v.label -i rh.wang2015atlas.V1d.label -i rh.wang2015atlas.V1v.label -i rh.wang2015atlas.V2d.label -i rh.wang2015atlas.V2v.label -o lhrh.wang2015atlas.V1d+v+V2d+v.label
# GL is the project name. funcroi-config -label lhrh.wang2015atlas.V1d+v+V2d+v.label -analysis GLanalysis.rh -roi rhlh.wang2015atlas.V1d+v+V2d+v.roicfg
funcroi-sess -sf GLsublist -roi rhlh.wang2015atlas.V1d+v+V2d+v.roicfg
funcroi-table-sess command showed an error "ERROR: there is a vertex in the label that cannot be matched to the surface. This usually occurs when the label and surface are from different subjects or hemispheres or the surface has been changed since the label was created."
The command showed no error with -anlysis GLanlaysis.lh option. Still, it looked a bit weird as you can see the attached file.
The attached files are the output from funcroi-sess with different hemispheres options.
What should I do for doing ROI analysis from the both hemispheres at once? If I can't, would it be good idea that I get mean beta values from each hemisphere, add them together and divide by 2 to get the average beta value?
Kind Regards, Andy (Junghyun) Kim