The anticipated impact of this study is defining voice biomarker features and reward functions for a deep reinforcement learning based system from group interactions that improve depression and anxiety outcomes. The investigators' ability to quantify the real-time impact of human-intervention in scaled group video sessions can be very meaningful for creating best practices in the area where measurement is infrequent.
The investigators' priority is to scale the optimal mix of individuals for group therapy sessions based on reward functions that maximize improvements in depression and anxiety scores. Current group therapy appointments may track little save few who use various group feedback questionnaires (e.g. OQ, GCQ, or GQ). Voice biomarkers can play a key role in the real-time measurement of mental health.
The proposed work is to conduct a feasibility study on creating reward functions that most effectively enable engagement for group sessions as measured by voice biomarkers before, during, and after group video meetings.
