American Diabetes Association presentation highlights innovations in technology for diabetic eye condition

Estimating the risk of DR progression is clinically difficult due to the task drawing upon medical knowledge and clinical experience that may vary between clinicians. (Adobe Stock image)

Findings from a study highlighting the latest developments in progression risk estimation for diabetic retinopathy was highlighted today as a late-breaking poster at the 83rd Scientific Sessions held by the American Diabetes Association (ADA) in San Diego, California.

Paolo S. Silva, MD, co-chief of Telemedicine, Beetham Eye Institute, Joslin Diabetes Center. associate professor of Ophthalmology, Harvard Medical School., presented his late breaking poster titled Identifying the Risk of Diabetic Retinopathy Progression Using Machine Learning on Ultrawide Field Retinal Images, during the General Poster Session on Saturday.

It is estimated that by 2050, the number of individuals with diabetic retinopathy (DR) could nearly double to impact more than 14 million Americans. Estimating the risk of DR progression is clinically difficult due to the task drawing upon medical knowledge and clinical experience that may vary between clinicians. The current DR severity scales inform clinicians on the progression risk providing recommendations for follow-up and treatment. This study sought to evaluate how the use of AI algorithms may improve this process.

The study, titled “Identifying the Risk of Diabetic Retinopathy Progression Using Machine Learning on Ultrawide Field Retinal Images,” examined the use of AI algorithms to improve the process of estimating the risk of DR progression. In this study, the authors developed and validated machine learning (ML) models for DR progression from ultrawide field (UWF) retinal images, which were labeled for baseline DR severity and progression.

Findings show the AI prediction for 91% of the images were either correct labels or were the labels with greater progression than the original labels. These findings demonstrate the accuracy and feasibility of using machine learning models for identifying DR progression developed using UWF images.

"Currently, estimating the risk of DR progression is one of the most important, yet difficult tasks for physicians when caring for patients with diabetic eye disease," Silva said. “Our findings show that potentially, the use of machine learning algorithms may further refine the risk of disease progression and personalize screening intervals for patients, possibly reducing costs and improving vision-related outcomes.