The Impact of AI on Diagnosis and Treatment of Eating Disorders :Advancing Patient-Centered Care

Abstract

Eating Disorders (EDs) are complex psychiatric conditions that often go underdiagnosed and undertreated. Artificial Intelligence (AI) offers new opportunities to enhance early detection, personalize treatment, and improve access to care. This narrative review synthesizes recent literature on the use of AI in the diagnosis and treatment of EDs, focusing on its potential to support patient-centered approaches. Studies show that machine learning models can accurately predict ED risk using data from self-reports, neuroimaging, and social media. Chatbots using natural language processing have demonstrated efficacy in delivering psychoeducation and reducing ED-related concerns, particularly among high-risk populations. AI tools are also being adopted by clinicians for administrative tasks and clinical decision support, though their impact on therapeutic outcomes remains limited. Despite these promising developments, challenges persist. Concerns about algorithmic bias, data privacy, lack of transparency, and limited real-world validation highlight the need for caution. Ethical questions surrounding the use of AI in emotionally sensitive care further complicate its integration. While AI shows strong potential to complement traditional ED care, its success depends on careful implementation, interdisciplinary collaboration, and ongoing research. When developed responsibly, AI can become a valuable tool in advancing equitable, effective, and patient-centered eating disorder services.