AI predicts heart attack

Changes in the small blood vessels of the retina are indicators of many cardiovascular diseases. In a new study organized by the University of Leeds, machine learning methods were used to prepare an artificial intelligence system for automatic analysis of retinal images and identification of patients who are likely to have an acute myocardial infarction within the next year.

Machine learning is a complex series of algorithms that allow computers to identify patterns in data and make predictions. Retinal photoregistration is a relatively inexpensive and often performed in the practice of an ophthalmologist. As a result of automated screening of images, artificial intelligence identifies patients at high risk of acute circulatory disorders, who can then be referred to specialized cardiology services.

In the process of machine learning, the artificial intelligence system analyzed retinal images and heart examination data from more than 5,000 people, revealing associations between retinal pathology and changes in the heart, including the size and function of the left ventricle. In addition, the data of each patient was supplemented with information about the age and gender of the patient.

After machine learning, an artificial intelligence system based on a retinal image and demographic data is able to make a forecast about the risk of myocardial infarction over the next 12 months with an accuracy of 70% to 80%. The system can also be used to track early signs of cardiovascular disease.

Thus, preventive measures and treatment can be initiated in a timely manner, which will significantly reduce cardiovascular mortality.

Article A.Diaz-Pinto et al. Predicting myocardial infarction through retinal scans and minimal personal information is published in the journal Nature Machine Intelligence.

Aminat Adzhieva, portal "Eternal Youth" http://vechnayamolodost.ru based on the materials of the University of Leeds: AI can identify heart disease from an eye scan.