Congratulations to the FreeSurfer team in the continued progress made in the FS tool set.  I wondered if you could comment on whether it is feasible to create a custom reference region (other than specifying, for example, cerebellar cortex or another predefined FS mask) for PETsurfer kinetic modeling (e.g. a cerebellum subfield) or whether there perhaps is an existing cerebellar subfield atlas that can be run with PETSurfer (e.g. crus2)?  In addition, can you please confirm whether the Logan model can be run using a reference region (i.e. without requiring arterial input)?  Thanks very much and kind regards.