AI-Enhanced Measure of Nearsightedness Improves Prediction of Retinal Damage Risk

A new method combining routine eye scans and artificial intelligence (AI) has been developed to measure nearsightedness (myopia) more accurately, potentially improving the prediction of long-term eye damage, according to a recent study published in Investigative Ophthalmology & Visual Science. This innovative approach, known as fundus refraction offset (FRO), takes into account subtle retinal changes that traditional glasses prescriptions often miss. The study, conducted by researchers from the University of Edinburgh, analyzed health data and retinal images from over 9,300 UK Biobank participants aged 40 to 69, none of whom had previous retinal issues. By using AI to compare the structural appearance of a person's retina to what is expected based on their glasses prescription, the researchers calculated an FRO score. They found that for each step lower in the FRO score—indicating a more nearsighted-looking retina than the prescription suggested—the risk of developing retinal detachment or tears over a 12-year period increased by about 50%. This finding is significant because while people with strong nearsighted prescriptions are generally 13 times more likely to experience retinal detachment compared to those without nearsightedness, the risk can vary widely among individuals with similar prescriptions. The FRO score provides a more nuanced and accurate assessment, helping to identify those who are particularly vulnerable to retinal complications. The potential applications of FRO are far-reaching. Clinicians can use this measure to better tailor their management strategies for myopic patients, potentially preventing vision loss. Additionally, FRO could play a crucial role in guiding decisions for elective procedures like laser eye surgery and clear lens exchange, where retinal vulnerability is a critical factor to consider, even if patients have the same glasses prescriptions. Fabian Yii, the principal investigator from the University of Edinburgh's Robert O. Curle Ophthalmology Suite in the Institute of Regeneration and Repair, emphasized the importance of this research. "Our study supports moving beyond glasses prescription alone and taking into account retinal information when describing myopia severity," he said. "Given the limitations of current descriptors of myopia severity, most notably glasses prescription, in effectively capturing an individual's risks of myopic complications, this research represents an important step in allowing care to be targeted more effectively for the increasing number of people affected by myopia." Industry experts and ophthalmologists highlight that the FRO score could significantly enhance the precision of myopia risk assessments, addressing a critical gap in current clinical practices. With myopia expected to affect nearly 50% of the global population by 2050, this AI-driven tool offers a promising solution for better managing this common eye condition and reducing the risk of serious vision problems. The University of Edinburgh's Institute of Regeneration and Repair is renowned for its cutting-edge research in ophthalmology and regenerative medicine. The institute's collaboration with AI technologies underscores a growing trend in healthcare, where advanced computational methods are being integrated to improve diagnostic accuracy and patient outcomes. Meta and Amazon are among the leading tech companies investing in such health-tech innovations to drive medical research forward. In summary, the FRO score, powered by AI, represents a significant advancement in myopia management, offering a more personalized and effective way to predict and prevent serious retinal complications. This could have profound implications for both clinical practice and the broader field of ophthalmology.