The medical sector is increasingly relying on artificial intelligence (AI) to interpret imaging Some of the most reliable indicators the algorithm found are highlighted, where red indicates elevated activity and blue indicates suppressed activity in the brain. Source: Sunil Kalmady, University of Albertaresults and diagnose diseases. A new system developed by researchers from the University of Alberta and India’s National Institute of Mental Health and Neuro Sciences diagnoses schizophrenia from patient brain scans, a determination traditionally based on subjective data of patient experiences.

The EMPaSchiz (Ensemble algorithm with Multiple Parcellations for Schizophrenia prediction) model has been trained on resting-state brain patterns using functional magnetic resonance imaging scans from many patients diagnosed with schizophrenia. The system is one of the first machine learning tools trained exclusively on data from patients who are not yet undergoing drug treatment, which could make it more valuable in the early stages of diagnosing the illness. EMPaSchiz identified the disease in new scans with 87% accuracy, outperforming existing AI models in diagnosing the disease.

The diagnostic method relies on a single modality of data acquisition for neuroimaging and its dependence on a set of predefined atlases renders it easily scalable. A research paper on EMPaSchiz is published in NPJ Schizophrenia.