What if the state of health was visible in the eye?

Artificial intelligence is turning the medical field upside down. Among these revolutions there is one that concerns the eyes. An AI, in fact, would make it possible to read the state of health through the latter.

Thanks to this technology, it will be possible to screen for Alzheimer’s disease and the risk of heart attack through retinal images. The availability of this artificial intelligence? Count a few years! A first report from the Hong Kong Medical School and experts from the English University of St George are more than positive about this revolutionary AI. By simply photographing the retina, we could therefore detect serious diseases. Learn about this artificial intelligence in more detail.

The retina, a gold mine for AI

A few months ago, another artificial intelligence was designed in the UK to determine a person’s remaining years to live by studying their eye. Fundus images of 47,000 individuals were analyzed to analyze the link between retinal age and mortality. Therefore, if the algorithm announces that a patient’s retina is one year older than its actual age, the occurrence of death, regardless of the reason, increases by 2% over the next eleven years. Ditto for the danger of death from a cause other than cardiovascular problems or cancer: there is a 3% increase.

According to their research, cells in the back of the eye could predict the onset of specific cardiovascular and kidney problems. The retina is an organ that can be easily studied by AI. By pushing the work, experts could anticipate disease and pinpoint the age of death.

Reading the state of health in the eyes: more reliable than today’s diagnostic techniques?

The first report on the subject is the work of the Hong Kong Medical School. Its lead author is Carol Cheung. This AI can find Alzheimer’s disease, via retinal images. We already knew that this pathology was mixed with deformations of this membrane present in the back of the eye. Therefore, they fed their algorithm thousands of images of the retinas of individuals with Alzheimer’s. Conclusion: the latter manages to make the diagnosis for over 80% of situations.

Often, to detect this disease, it is necessary to mix memory tests, difficult samples and even a CT scan. A Difficult Course On the other hand, if you have been to an ophthalmologist, you are undoubtedly familiar with the type of equipment a retinograph is. In a matter of minutes you get a snapshot of the retina. Then the photos will be shown to the famous AI of Carol Cheung, to have a diagnosis of the state of the retina … and also of the brain! However, the expert said it is still necessary to validate the findings with individuals in the early stages of Alzheimer’s.

Through these photos, it is possible to identify a danger of heart attack or stroke

The team of Professor Alicja Rudnicka, of the English University of St George based in the capital London, found that very small vessels of the retina were the mirror of the state at time T of the other vessels of the organization. So, if the latter remains too narrow and winding, this is a bad omen. Therefore, these experts have conceived and designed an AI, using retinal images of over 88,000 volunteer individuals, benefiting from a follow-up for years, in the British Biobank. It is a large-scale biomedical research and database. They integrated known risk factors such as smoking or hypertension.

And it works. Their artificial intelligence is able to find the individuals most at risk of heart attack or stroke and the main causes of death linked to a degraded state of the arteries. Since this is a public type search, the algorithm is free. Professor Alicja Rudnicka does not want the project to spread over time. Therefore, you want to design a reliable and secure digital link between the retinography screening network and that of general practitioners who will care for people on “red alert”.

Leave a Reply

Your email address will not be published. Required fields are marked *