Hi Elizabeth, when FS computes a p-value, it computes assuming an unsigned test. When you specify a sign to mri_surfcluster (eg, --sign pos), it recomputes the p-value to take advantage of the signed test (divides the p-value by 2, same as lowing the threshold from 2 to 1.7. So try using 1.7 instead of 2 in tksurfer. Also, why are you setting a max for surfcluster? That will exclude anything above 5.

doug

Elizabeth Gutierrez wrote:

Hi all,

We're trying to create labels from a group level sig.mgh file. However, after the labels are created and an annotation file/color table is made, we try to view the sig.mgh file and the annotation file in tksurfer, but they don't extactly overlap, even though they are both configured to the same threshold (min-2 max-5). Most of the labels are fine, but the labels closest to the the higher end of the threshold end up being hollow, and only outline the sig.mgh clusters instead of being completely filled out. Does that make sense?

Here is our command for mri_surfcluster:

mri_surfcluster --in sig.mgh --hemi lh --subject fsaverage --thmin 2 --thmax 5 --sign pos --no-adjust --cwsig --o ./output_id --ocn ./cluster_number --olab ./grouplabel

Thanks in advance! Elizabeth

-- Elizabeth C. Gutierrez Graduate Student Department of Brain and Cognitive Sciences Massachusetts Institute of Technology

---

---

Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer