An algorithm makes it possible to analyze the fundus of the eye, and to identify abnormalities in the blood vessels, characteristic of atherosclerosis.
- Atherosclerosis corresponds to the formation of plaques on the walls of the arteries, composed of lipids.
- It is the leading cause of heart attacks and strokes.
- A fundus examination allows you to observe the blood vessels, and to detect the early signs of the disease.
Our eyes allow us to see, but they could also become an indicator of cardiovascular pathology. German researchers have developed a method to detect atherosclerosis, a disease of the arteries, using images of the eyes. In their study, published in the journal Scientific Reportsthey explain that it is the blood vessels, very numerous in the eyes, which make it possible to identify the signs of the disease.
Why use the eye to screen for atherosclerosis?
“As it usually causes no symptoms for the first few years, the diagnosis is often not made until side effects have already occurred.“, explains Dr. Nadjib Schahab, one of the authors of the study. However, the pathology causes circulatory problems in the legs and arms, which can lead to amputation in the most serious cases. “The risk of a fatal heart attack or stroke is greatly increased – even in the early stages of the disease“, adds the scientist. With his team, he is working on an early detection method to prevent the pathology from progressing to these stages. Now arteries and veins can be easily observed and photographed through the pupil.
An algorithm trained to recognize pathology
To test their hypothesis, the study authors photographed the eyes of 97 women and men who suffered from atherosclerosis. “In more than half of them, the disease was still at a stage where it caused no symptoms. says co-author Dr. Maximilian Wintergerst. At the same time, images were taken of 34 people without the disease. In a second step, the images were scanned in a software, called convolutional neural network: it is inspired by the human brain in its functioning. The scientists taught the software to recognize signs of the disease, first training it with another eye vessel pathology. This allowed the algorithm to detect whether the presented eye imagery was from a patient with atherosclerosis or not. “80% of all presented individuals were correctly identified, taking into account 20% false positives, i.e. healthy individuals that the algorithm incorrectly classified as diseasedsays Thomas Schultz, another co-author. This is amazing, because even for trained ophthalmologists, atherosclerosis cannot be detected from fundus images..”
A future screening technique?
The authors of the study do not stop at these first exciting results: they wish to further improve their method. In particular, they plan to cooperate with ophthalmology and vascular medicine centers to obtain additional fundus images of people affected by atherosclerosis. In the long term, their goal is to develop a simple, fast and reliable diagnostic tool.
.