Can someone please point me to a detailed description of how FSL
implements its prewhitening (FILM) step. In particular, is a global
estimate of AR(1) derived and then applied similarly to all voxels? i.e.
one pre-whitened design matrix for all voxels.
Thanks.
|