Hi,

I am sorry for insisting, but I have found a new mistake in your new version of make_average_surface.

rh.medial_wall.label contains 13733 vertex, it is a desired result. However, rh.cortex.label contains all 163842 vertex, i.e, it covers medial wall area too. That's why we still have the same problem because we want to do statistic tests just over the cortex (without medial wall). It also happens with left hemisphere.

Could you please take a look at that issue?

Thanks very much,

M del Mar