We used to use a structural T1 image co-registered to the functional
images, and register this to a T1 template image. However, we found
that we got slightly better results if we used a mean of our functional
images and registered this to a template of the same modality. This was
especially the case with fMRI, since the rigid body registration between
the structural and functional images was never particularly good because
of the distortions in the functional images.
Regards,
-John
> can anybody tell me which image to choose for the best in
> normalisation procedure when the SPM asks:
>
> "image to determine parameters from?"
>
> Thank you very much!
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|