I would like to vertex-wise compare the left and right hemisphere thickness in a group of subjects. 

Would it be okay to just substract lh.hickness.0.mgh and rh.thickness.0.mgh (i.e. the files concatenating the thicknesses of all patients), then smooth and then perform a one sample t test? Or does freesurfer include a more sophisticated test to perform such a comparison?

Best,

Martijn