Hi,
I'm new to Freesurfer and am still figuring out the ins and outs of cortical reconstructions. One option I've been looking into is the "Smooth Surface" tool in Freeview. It seems to make it easier to identify handles and other errors especially in subjects with significant atrophy, but I'm not sure what's the best way to use it.
I did a few trial runs with various smoothing parameters, and it looks like running autorecon2-wm and autorecon2-pial with autorecon3 both eliminate the effects of the smoothing (parcellation stats don't change), while running autorecon3 alone changes all the stats other than the number of vertices and thickness averages. In particular, the surface area and gray matter volume decrease by 5-10% depending on the parameters I choose (taubin vs. standard, number of iterations, lambda/frequency cutoff), which isn't unexpected. Would you recommend using smoothing to help fix small errors and if so, what parameters would you recommend?
Thanks, -Caleb
freesurfer@nmr.mgh.harvard.edu