I have a general question about handing atlas ROIs after applying recon-all on a T1W image. Could you please clarify how I can minimize/prevent atlas ROIs overlap in volumetric subject-specific space when going from surface-based space to volumetric space?