External Email - Use Caution

Hello FreeSurfer Developers,

I run recon-all on the T1 and T2 images of a few new human subjects. After running recon-all, I followed the freesurfer troubleshooting tutorial, and started visually checking the segmentation of the pial surfaces and white surfaces. I found similar errors among a few subjects, and would like to make sure that what I am thinking is correct before proceeding: The figures I attached are the coronal view of the same subject (from the left column to the right column: posterior to anterior). These errors mainly occurred at the occipital cortex, and at the boundary between the left and right hemisphere. It seems that the pial surface included a piece of dura matter that should not be there. It is indicated by the green arrows in the images. It also seems to happen at the places indicated by the blue arrows, but not as visible. Is what I'm thinking correct? I also attached an image containing three different views of the same subject. In general, will running -gcut in this example help reducing the errors above?

-The recon-all is run on a server with freesurfer-Linux-centos6_x86_64-stable-pub-v6.0.0-2beb96c -uname -m x86_64 -recon-all.log: see attached

Thank you so much.

HHL