Artificial intelligence can predict health risks through your retina

  • News
  • Published on

    Reading 3 mins.

    According to British research, AI-enabled imaging of the network of veins and arteries in the retina could accurately predict cardiovascular disease and even death.

    The announcement may seem futuristic, yet it dates from October 4. According to a study involving several English scientific and medical universities, AI-enabled imaging of the network of veins and arteries present in the retina could now accurately predict the cardiovascular diseases that await you, and even possible death according to your health status. This without the need for blood tests or blood pressure measurement, making the measurement much less invasive, so cardiovascular disease, heart failure and stroke are leading causes of ill health and death in the world. world. The findings of this study are published in the British Journal of Ophthalmology.

    An algorithm created to detect known risk factors

    The retina as an indicator of health status is not new. Previously published research indicated that the width of tiny veins and arteries (arterioles and venules) in the retina can provide an early and accurate indicator of circulatory disease.

    The British researchers therefore had the idea of ​​developing a fully automated algorithm enabled by artificial intelligence (AI) (QUantitative analysis of the topology and size of the retinal vessels, or QUARTZ in short English) to develop models allowing to Assess retinal vascular imaging, as well as known risk factors.

    They applied QUARTZ to retinal images of 88,052 UK Biobank participants aged 40 to 69, looking specifically at the width, vessel area and degree of curvature (tortuosity) of retinal arterioles and venules. They then applied these models to retinal images of 7411 participants, aged 48 to 92, from the European Prospective Investigation into Cancer (EPIC)-Norfolk study. QUARTZ performance was also compared to Framingham risk scores.

    The health of all participants was followed for an average of 7-9 years, during which time there were 327 deaths from circulatory disease among participants in the UK Biobank and 201 circulatory deaths among 5862 participants in those of EPIC- Norfolk

    In men, arteriolar and venular width, tortuosity and variation in width emerged as important predictors of death from circulatory disease. In women, arteriolar and venular area and width and venular tortuosity and width variation contributed to the prediction of risk. Overall, these predictive models, based on age, smoking status, medical history and retinal vasculature, captured between half and two-thirds of circulatory disease deaths in those most at risk. A score almost on par with the Framingham risk score.

    A non-invasive practice for better prevention?

    Retinal imaging is already common practice in the UK. “AI-based vasculometry risk prediction is fully automated, inexpensive, noninvasive, and has the potential to reach a greater proportion of the population.” lists the study. The researchers see it as a predictive tool of choice to invest more: “The retinal vasculature is a microvascular marker, so it offers a better prediction of circulatory mortality and stroke.”

    Above all, the non-invasive form of AI imaging would make it possible to reach a wide audience, without even having to take a blood test. The study affirms that in the general population, it could be used as a non-contact form of systemic vascular health check-up, to triage those at medium to high risk of circulatory mortality for further clinical risk assessment and appropriate intervention. And could be included in the health check for people aged 41 to 74.

    “What is needed now is for ophthalmologists, cardiologists, primary care physicians and computer scientists to work together to design studies to determine if using this information improves clinical outcomes,” concludes the study.

    We want to say thanks to the author of this article for this awesome content

    Artificial intelligence can predict health risks through your retina

    Our social media profiles here as well as additional related pages here.