Multimodal Sensing of Therapist-Client Synchrony to Predict Psychotherapy Outcomes

We propose using wearable sensors, facial expression recognition, and speech processing to measure physiological, nonverbal, and linguistic synchrony between therapists and clients during Motivational Interviewing sessions for heavy drinking. By analyzing these multimodal signals, the project aims to predict metrics of clinical experiences and treatment outcomes, enhance theories of therapeutic interaction, and develop real-time digital biomarkers for clinical feedback.

clinical experiences and treatment outcomes of psychotherapy

multimodal sensing technologies and data modeling techniques

interpersonal coordination in dyadic interaction