Integrating Artificial Intelligence and Smartphone Technology to Enhance Personalized Assessment and Treatment for Eating Disorders

OBJECTIVE: Smartphone technology presents a promising path toward expanding access to evidence-based eating disorder assessment and treatment. Despite rapid technological advances, research has yet to harness these systems in ways that make personalized digital health care a clinical reality. In this forum, we review extant research testing smartphone intervention and monitoring tools for eating disorders and explore innovative ways integrating this technology with AI can enhance assessment, symptom detection, and intervention efforts. METHOD: We highlight three capabilities of smartphones that hold promise for delivering personalized and maximally effective digital health tools: (1) passive sensing and digital phenotyping; (2) natural language processing of reflections from in-app homework tasks; and (3) closed-loop adaptive interventions. We discuss how these capabilities can augment current assessment and treatment efforts and draw on literature from other fields to inform research questions for the eating disorder field. RESULTS: Evidence from other fields demonstrates the feasibility of constructing data-driven models from smartphone sensor data and textual input from in-app CBT activities to predict clinical outcomes. These models may inform closed-loop interventions, enabling apps to deliver timely, personalized support in response to real-time changes in a user's needs. CONCLUSION: The eating disorder field can draw on lessons from other fields to evaluate smartphone technology that leverages AI to enhance personalization. Realizing the potential of these tools will require addressing challenges related to engagement, trust, data governance, and clinical integration. The testable research questions presented here offer a roadmap to guide future large-scale, collaborative efforts aimed at transforming eating disorder care.