Home LATEST NEWS Algorithm detects heart attack risk based on eye exam

Algorithm detects heart attack risk based on eye exam


Work by Prof Alex Frangi and his colleagues at the University of Leeds in the UK shows that an artificial intelligence (AI) algorithm is able to spot small changes in retinal blood vessels indicative of risk high rate of myocardial infarction with an accuracy of between 70% and 80% among the 5000 study participants.

Deep learning (PA) is a complex series of algorithms that allow computers to identify patterns in a dataset and make predictions. Once the image patterns are learned, the PA estimates the size and pumping efficiency of the left ventricle, one of the heart’s four chambers, from retinal scans. You should know that an enlarged ventricle is linked to an increased risk of heart disease.

It is through this information about the left ventricle, combined with basic patient demographics (age and gender), that the algorithm predicts heart attack risk over the next 12 months.

Read:  Extreme heat: Canada must prepare for deadly episodes, says report

Analyzing Retinal CT Scans Using AI Could Revolutionize Heart Disease Screening, explains the university in a press release.

From ophthalmologist to cardiologist

According to these researchers, AI could thus be used to refer a patient with clues in the retina to a cardiologist who can perform a thorough cardiovascular examination.

Our deep learning program has the potential to identify those at higher risk of cardiovascular disease among people undergoing routine eye examinations, allowing preventive treatments to be started earlier and avoid premature cardiovascular disease, notes Chris Gale, professor of cardiovascular medicine at the University of Leeds and one of the authors of the work published in the journal nature machine intelligenceHave (New window)Have (in English).

Retinal CT scans are relatively inexpensive and are routinely used in many eye clinics. They are used to diagnose and monitor primary eye diseases, such as diabetic retinopathy.

Read:  The urgency of supervising cyberathletes as athletes in their own right

Currently, the size and pumping efficiency of a person’s left ventricle can only be determined if they have heart diagnostic tests such as echocardiography or magnetic resonance imaging. These tests can be expensive and are often only available in hospital settings, making them inaccessible to many people who live in countries with less well-resourced healthcare systems.

Other studies have shown in recent years that deep learning algorithms are able to identify certain cancers such as skin cancers on photographs or breast cancers from mammography images.

Previous articleKurds regain control of prison attacked by IS | Syria: the spiral of war
Next articleCOVID tests are testing the patience of Beijingers