Hi Aaron,
This is a problem that occurs when you have high spatial frequency and too few iterations for the LB_FEM method. For another picture of the artifact see: [
afni.nimh.nih.gov]
Also, if you like to see it grow, run the program with the -talk_suma flag.
At any rate, increase the number of iterations and this will go away. You can also use the HEAT method with about 100 iterations too. That method will not cause such artifacts.
Note that I am in the process of revamping SurfSmooth. It turns out the HEAT's approximation of the FWHM can be off. But I won't have these fixes ready for a while.
cheers,
-z