Dear Dr. Friston:
Thank you for your reply. Now, we have a better understanding of what's
going on.
My next set of questions concerns smoothing: I have read somewhere on the
list that the size of the Gaussian kernel should be 2-3 times the voxel
size of the normalized image. But our voxel sizes are 2x2x4 mm (x, y, z).
We have been using 12 mm kernel so far, but now, we are questioning that.
Is smoothing function implemented isotropically in SPM'96 (that is, is
smoothing the same in all directions)? If so, is it possible that we are
oversmoothing our data?
Many thanks,
Vitaly Furman
Brain Imaging Lab
Columbia-Presbyterian Medical Center
New York, NY
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|