AI system uses retinal scans to predict heart attack risk

By analysing eye scans taken during a routine visit to an optician, the system can identify patients and predict with 70 to 80 per cent accuracy whether they are at risk of a heart attack over the next year.

This is because changes to the tiny blood vessels in the retina are indicators of broader vascular disease, including heart problems.

It's a breakthrough - reported in Nature Machine Intelligence - that could transform the way we screen for and track the early signs of heart disease, according to Professor Alex Frangi, who holds a Diamond Jubilee Chair in Computational Medicine at the University of Leeds, UK, and is a Turing Fellow at the Alan Turing Institute.

"Cardiovascular diseases, including heart attacks, are the leading cause of early death worldwide and the second-largest killer in the UK," Frangi says. "This causes chronic ill-health and misery worldwide."

"This AI system has the potential to identify individuals attending routine eye screening who are at higher future risk of cardiovascular disease, whereby preventative treatments could be started earlier to prevent premature cardiovascular disease," says Chris Gale, professor of cardiovascular medicine at the University of Leeds.

The researchers used deep-learning techniques - a complex series of algorithms used to enable computers to identify patterns in data and then make predictions - to train the AI system.

During this process, the AI system analysed retinal and cardiac scans from more than 5000 people, using data from the UK Biobank, to identify links between the pathology of the patient's retina and changes in their heart.

Having learned these image patterns, it could then estimate the size and pumping efficiency of the left ventricle (one of the heart's four chambers) from retinal scans alone. An enlarged ventricle is linked with an increased risk of heart disease.

Taking this estimated size and pumping efficiency of the left ventricle and combining it with basic demographic data about the patient (their age and sex), the AI system could then make a prediction about their risk of a heart attack over the next 12 months.

This is significant as the size and pumping efficiency details of the left ventricle can currently only be determined through expensive diagnostic tests, such as echocardiography and magnetic resonance imaging (MRI) of the heart.

These diagnostic tests are often only available in a hospital setting, which makes them inaccessible to people from countries with less well-resourced healthcare systems, and can increase healthcare costs and waiting times in developed countries. This research shows that there may be an easier, less expensive way to screen for cardiac disease.

"This technique opens-up the possibility of revolutionising the screening of cardiac disease. Retinal scans are comparatively cheap and routinely used in many opticians' practices. As a result of automated screening, patients who are at high risk of becoming ill could be referred to specialist cardiac services," Frangi says.

"The AI system is an excellent tool for unravelling the complex patterns that exist in nature, and that is what we have found here - the intricate pattern of changes in the retina linked to changes in the heart," adds Sven Plein, British Heart Foundation professor of cardiovascular imaging at the University of Leeds.

This article is published in partnership with Cosmos Magazine. Cosmos is produced by The Royal Institution of Australia to inspire curiosity in the world of science.