Hi, thank you for the great work! Quick question: does your model support multi-modal inputs like CT and PET scans? If so, how does it handle these different image types? Thank you!