Researchers at Shenzhen University have developed an artificial intelligence model capable of identifying individuals at risk of depression years before clinical diagnosis, potentially transforming early intervention strategies.
Chinese researchers say they have built an AI system that could flag a person’s depression risk as much as four years before diagnosis, a development they argue could open the door to earlier intervention and prevention. The team at Shenzhen University says the work is centred on major depressive disorder, a condition that affects more than 332 million people worldwide and remains difficult to treat, according to the South China Morning Post.
The model was trained on data from two European clinical studies of adolescent depression. In those trials, participants were assessed at ages 16, 19 and 23 through MRI scans, blood tests and questionnaires to determine whether they developed depression later on. The Chinese researchers analysed that information and used it to build an artificial intelligence model designed to identify risk patterns before symptoms became clinically clear, the South China Morning Post reported.
The approach fits into a broader push to use machine learning on health records and biological data to spot disease earlier. Researchers at the European Molecular Biology Laboratory have already developed Delphi-2M, a generative AI model that can estimate the likelihood of more than 1,000 diseases years ahead using large anonymised health datasets, according to reports on the study. That work, along with similar efforts in psychiatry, points to growing confidence that AI may be able to find warning signs in complex medical data that are hard for clinicians to see unaided.
But the field also faces important limits. The US National Institute of Mental Health has said AI tools based on smartphone behaviour may not reliably predict depression risk across diverse groups, after research found that such data can be uneven and hard to generalise. That caution suggests the Shenzhen model, while promising, will still need rigorous external testing before it can be judged useful in routine care.
Disclaimer: This content is for informational purposes only and is not intended to be a substitute for professional medical judgment, advice, diagnosis, or treatment.
