Using Artificial Intelligence to Detect Relapse of Psychosis: A Scoping Review (Preprint)

BACKGROUND Psychotic disorder represents a leading cause of disability worldwide, and relapse in psychosis is common. Artificial intelligence (AI) is increasingly recognized as a method which could aid clinical monitoring in psychosis. OBJECTIVE This scoping review aims to identify studies which have used methods with an AI component to detect relapse in psychosis. METHODS A systematic search strategy was conducted on PubMed, PsycINFO and Embase from inception to January 2026. Observational studies, randomized controlled trials and quasi-experimental studies which used AI methods to detect relapse in psychosis were eligible for inclusion. Screening and data extraction procedures were conducted by at least two reviewers working independently. Findings were extracted, charted and described using narrative synthesis based on data extraction and consensus meetings with the research team. The scoping review was prospectively registered with Open Science Framework. RESULTS Relevant studies identified (n = 10) included use of digital tools such as smartphone and smartwatch-based monitoring, ecological momentary assessment tools, social media activity and internet searches. Digital phenotyping via smartphones and wearables emerged as the most common method for data collection. Efficacy of AI models varied with sensitivity (or recall) ranging from 0.25 to 0.77 and specificity ranging from 0.06 to 0.88. Reported area under the receiver operating characteristic curve for models ranged from 0.63 to 0.78. AI models were heterogenous across studies, and most study findings were not replicated. CONCLUSIONS This scoping review highlights both the promise and current limitations of AI in psychosis relapse prediction. Digital phenotyping research in detection of psychosis relapse has progressed, but future studies need to include larger numbers of participants and should incorporate other methods such as use of large language models. Future studies will require large collaborations aiming to deliver AI tools for use in real world clinical practice. CLINICALTRIAL N/A