Dear Ambra,
Why not just go with the default asymmetric tissue probability maps during segmentation. Your data is asymmetric, and obviously asymmetric priors should be better priors than symmetric ones. Based on the rc files obtained during segmentation and the flipped rc files you can then create a study-specific symmetric Dartel or Shoot template, which you can then additionally register onto one of the symmetric MNI templates if desired.
In the past, you had to use symmetric priors to end up with (more or less nicely matched) symmetric data and corresponding flow fields, but with Dartel and Shoot this is basically obsolete.
Best regards
Helmut
|