Hi Afni Experts,
For slow_surf_clustsim.py I am not sure if I need to supply a mask or not for cortical thickness analysis. I used the -on_surface yes flag as shown below. In doing so do I also need to enter in a mask of cortical thickness since it is a thickness or is this taken into account with the simulation? When using SurfFWHM on the thickness files (smoothed 15mm) the estimate was about 10mm smoothness. I used this as an estimate and my input was the fsaverage left hand spec
slow_surf_clustsim.py -save_script surf.sim -on_surface yes -uvar blur 10 -uvar spec_file fsaverage/SUMA/fsaverge_lh.spec -uvar surf_vol fsaverage/SUMA/fsaverage_SurfVol+orig
The athresh of 1.64 was the max value reported by slow_surf_clustsim in the terminal for a p of 0.01 and it is below the freesurfer estimate of log(0.01)=2. Is the reason for the discrepancy based on the blur level estimate for why it is below the expect T threshold of 2? For p of 0.05 the log(0.05)=1.3 matches up very close to the 1.28 output from the terminal of slow_surf.
Additionally on SurfClust I used the following:
SurfClust -i ../SUMA/lh.pial.asc -input 'Lthickness.niml.dset[7]' 0 -rmm -2 -out_roidset -prefix HCvsGroupA_Rthickness_05 -amm2 463 -athresh 1.64
Since I am measuring thickness does the option of -i SUMA/lh.pial.asc make sense? I saw some other examples which use the SurfA pial option and I wanted to see which is correct (or if both get you to the same point).
Thanks,
Ajay
Edited 2 time(s). Last edit at 05/02/2016 11:55AM by AjaySK.
