A system has been developed that "reads" the risk of heart attack in the eyes

European and other scientists have developed an innovative artificial intelligence system that can analyze images from a human's eye and find out who is at increased risk for heart attack. The system has an accuracy of 70% to 80% in predicting a heart attack within the next year.

Researchers from Britain, France, Belgium, the United States and China, led by Alex Frangi, a professor of computer medicine at the University of Leeds and the Alan Turing Institute in Britain, published their findings in the journal Art Nature Machine Intelligence ", estimated that in the future there may be a revolution in the diagnosis of cardiovascular risks, especially for heart attack.

Doctors know that changes in the tiny blood vessels of the retina of the eye are indicators of wider vascular health or disease, thus revealing heart problems. The new development is that using "deep" machine learning techniques, an artificial intelligence system has been trained to automatically read retinal images and recognize in them the signs of a risk of heart attack.

Deep learning is a complex set of algorithms that allow computers to recognize patterns in data and make predictions. The new "smart" system analyzed retinal images and cardiovascular examinations from 5.000 people to learn how to correlate eye conditions with heart health.

See link for scientific publication HERE

Source: RES-EAP