AI diagnostics: Artificial intelligence to predict heart attacks

Researchers at the University of Leeds have worked with numerous partners to develop an AI system that can use eye scans to detect heart attack risk within the next 12 months. The AI diagnosis is based on the finding that changes in tiny blood vessels in the retina can indicate vascular disease. The AI system, developed under the leadership of the University of Leeds, can detect these changes on retinal scans and identify people who could suffer a heart attack within a year. The accuracy of the prediction system is between 70 and 80 percent, according to the researchers. It is therefore well suited as a recommendation mechanism for in-depth cardiovascular screening.