AI model identifies genetic patterns linked to ischemic stroke risk

Researchers have developed an artificial intelligence model using transformer technology to identify genetic patterns linked to ischemic stroke. Unlike previous tools that analyzed genetic differences individually, this AI model examines how multiple DNA variations work together across the genome.

By training on data from the UK Biobank and the US All of Us research program, the model identified 18 genetic variants associated with ischemic stroke. The study found that individuals in the highest genetic risk group were approximately 22% more likely to experience an ischemic stroke compared to those in the lowest risk group.

A key finding identified the EDEM2 gene in the brain's outer layer as a significant contributor to stroke susceptibility. Researchers at the University of Technology Sydney are now working to expand this into a multimodal AI tool that integrates genetic data with lifestyle surveys, medical scans, and imaging to improve personalized risk assessments.