Design and Formative Evaluation of a Voice-based Virtual Coach for Problem-Solving Treatment

Artificial intelligence (AI)-based voice technology offers considerable promise in healthcare; however, its application for behavioral therapy in a real-world or research setting has yet to be determined. We describe the design and evaluation of LumenTM, a fully functional voice-only virtual coach that delivers evidence-based, problem-solving treatment (PST) for patients with mild-to-moderate depression and/or anxiety. Participants (N=26) completed surveys and semi-structured interviews after two therapy sessions with Lumen. Participants found Lumen to provide high pragmatic usability and favorable user experience, with marginal task load during interactions. Participants emphasized its on-demand accessibility and the delivery of a complex PST task with a simplistic structure for achieving therapy goals. Suggestions for improvement included streamlining and improved personalization and pacing of conversations and providing additional context during therapy sessions. Lumen offers a realistic and cognitively plausible verbal interaction that can potentially lead to personalized and accessible mental health care, filling a gap in traditional mental health services.