Hi,

i want to extract epi time series on the surface, seems if i want every last drop of accuracy from bbregister i will need to run

bbregister --mov epi_i_.nii --bold \ --s fbirn-anat-101.v4 \ --init-fsl --reg register_i_.dat

on every epi volume in the time series and then mri_vol2surf --mov epi_i_.nii \ --reg register_i_.dat \ --projfrac 0.5 --interp nearest \ --hemi lh --o lh.epi_i_.mgh

or else if i just use the linear reg (flirt or something ) to the first epi and use the bbreg transform to the first for all the others i might as well just be doing linear registration, i.e. losing the power of bbreg on all the others.

then of course you have the problem of manually? checking all 300 or whatever registrations for the time series X number of subjects, or some auto method to check for time series outliers as a way to see what registrations may have gone of the track, but thats another story

or am i missing something ?

thanks

greg