Schizophrenia is a severe, chronic mental disorder affecting about 1% of the global population. Traditionally, diagnoses have relied primarily on behavioral assessments, and the medical community has long sought objective biological markers. Professor Chih-Chieh Yang from the Department of Psychiatry at Taipei Veterans General Hospital has led a team in developing an advanced “Smart Brain Imaging Platform” using cutting-edge medical AI technology. This platform assists in evaluating schizophrenia with an accuracy rate of 91.7%, earning a Gold Medal at the prestigious Edison Awards—known as the “Oscars of Innovation.”
Patients with schizophrenia often experience symptoms such as hallucinations, delusions, and declining social function. These are believed to stem from disruptions in neurotransmission and structural and functional degeneration of the brain. For example, a 30-year-old patient surnamed Chang was experiencing auditory hallucinations and paranoid delusions that interfered with daily life. The smart platform revealed abnormalities in the insular and temporal lobes, supporting a diagnosis of schizophrenia and enabling personalized treatment recommendations.
Yang explained that due to the lack of objective diagnostic tools for psychiatric disorders, medical assessments have typically relied on clinical observations and subjective interviews, which limit the ability to fully grasp disease progression and brain changes. Since 2012, the Smart Brain Imaging Platform has been developed using data from over 1,500 local participants. By integrating MRI scans with deep learning algorithms, the team successfully identified degrees of degeneration in regions like the insular cortex and temporal lobes—areas not easily visible to the naked eye—achieving a diagnostic accuracy of 91.7%.
Because schizophrenia is a chronic condition, patients often show significantly reduced function in the insular and temporal lobes compared to healthy individuals. These regions are crucial for sensory integration and emotional expression. Through AI-based computation, the platform can accurately pinpoint abnormal brain regions, aiding physicians in clinical interpretation.
Yang noted that the platform not only establishes predictive markers for brain aging and lesion progression but also offers 3D visualization capabilities to help doctors better interpret findings and communicate with patients and families. It also provides MRI-guided navigation, enabling the development of precise, non-invasive neurostimulation tools such as transcranial magnetic stimulation for therapeutic support.
According to Yang, the platform represents a major breakthrough by moving beyond traditional symptom-based assessments and offering objective biological indicators. It lays a new foundation for the diagnosis and personalized treatment of psychiatric disorders. To further support follow-up and therapeutic innovation, the research team has applied for a clinical trial with Taiwan’s Ministry of Health and Welfare to establish a new smart neuropsychiatric treatment center focused on developing more precise transcranial stimulation therapies for schizophrenia.
Resource: 北榮醫療AI診斷里程碑！腦影像揪思覺失調症 準確率達91.7%
