The answer will depend on your data, but for data from the FIL, we found that
the warps produced by SPM5's segmentation routine were more accurate than
those from the spatial normalisation. Jenny Crinion had a poster about this
at HBM.
Best regards,
-John
> Could you please tell whether SPM5 normalization algorithm is better than
> SPM2 from the point of view of the correspondence of cortical structures?
> If yes, how they are different?
|