I'm using data that's 0.35mm isotropic (and lower, but I've just noticed this). to do that I alter the header so that the voxels claim to be 1mm. it's a small brain so this works. Everything in general is fine.
the only thing is I'm having touble with the cortical thickness output of mris_make_surfaces. I can't remember if this is a new problem, or has just never mattered to me before.
when I run mris_make_surfaces, the ?h.thickness data seems to top out at 5. So that most of the brain has a thickness of 5.
I wonder does freesurfer top out the thickness metric data at 5mm, since humans (or monkeys) don't normally have GM that is more than 5mm thick?
if that's the case, can I change the behavior?
I hope it is my error, and the thickness is topping at 5 for some reason I can control.
attahced: pial surface with metric on it and scale bar sagittal slice with little measurements drawn to estimate thickness.
if you multiply the meaurements by 0.35 they are roughly what you'd expect.
but many are >5mm before scaling. they all report as 5mm.
I really need just anyway to calculate the thickness. that would be great. It needs to be quite accurate.
maybe I could change the volume back to 0.35 conformed, then scale the surfaces as well, and use mris-refine or something like that? the data would be the same but freesurfer would now know the true dimensions when it ran mris_refine. or anything that generates ?h.thickness.
many thanks,
Colin
