Abstract
Machine learning-based audiovisual phenotyping can reveal hidden discrepancies between patients' self-reported experiences and nonverbal expressions, offering a promising tool for objectively assessing communication quality and advancing health equity.