Hi Bin,
That’s correct, the -stop option will not accept an input image, rather, it flags to stop propagating a streamline once it has traversed all include regions.
If the thalamus region is labelled as SGM in the 5tt.mif image, then streamlines should terminate within the thalamus. As such, the stop option might be redundant.
Does the dt_l_local.tck output look sensible?
Cheers,
Arkiev
PS: This paper might be of interest for DRTT reconstruction…