Hi,
I'd like some clarification on the recommended procedure for smoothing functional data across the cortical surface in FS-FAST.
I've managed to get it working using the following procedure:
1) analyze my data using selxavg3-sess in the native space, without smoothing 2) resample the averages into the spherical space using func2sph-sess 3) smooth these averages using sphsmooth-sess (which required several edits to work with .nii files) 4) recreate the contrasts in the smoothed sph space with stxgrinder-sess 5) paint the analysis onto the surface with paint-sess (which also required edits for .niis) 6) view with surf-sess (which I had to edit to read the subject properly)
Surely there's a simpler way to accomplish this? I'm particularly confused because the online documentation says that selxavg3 replaces stxgrinder, but it seems stxgrinder has to be called to make contrasts in non-native space (and uses different code to produce the contrasts than selxavg). Plus, the revision date for stxgrinder is more recent!
Then, it seems odd that isxconcat will reproject the data for a group analysis if it's already sampled on the surface via func2sph.
Thanks, Scott