JAMA Psychiatry

AI predicts mental illness trajectory: How accurate is it?

February 26, 2025

Machine learning can feasibly detect schizophrenia progression using routine clinical data, potentially reducing diagnostic delays and untreated illness duration.

Study details: This cohort study utilized electronic health records from the Psychiatric Services of the Central Denmark Region to train machine learning models for predicting the transition to schizophrenia or bipolar disorder. The study included 24,449 patients aged 15 to 60 years with ≥2 psychiatric service contacts between 2013 and 2016. Predictors included medications, diagnoses, and clinical notes, and the main outcome was the diagnostic transition within 5 years.

Results: AI models predicted progression to schizophrenia with more accuracy than for bipolar disorder. For schizophrenia, sensitivity was 19.4%, specificity, 96.3%, and positive predictive value, 10.8%. Clinical notes--detailing symptoms, treatment responses, and patient-clinician interactions--were particularly valuable for prediction.

Source:

Hansen L, et al. (2025, February 19). JAMA Psychiatry. Predicting Diagnostic Progression to Schizophrenia or Bipolar Disorder via Machine Learning. https://pubmed.ncbi.nlm.nih.gov/39969874/