Evaluating AI Software to Help Detect Diabetic Retinopathy in Adults with Type 2 Diabetes

What this Trial Is About

Researchers are evaluating the use of an artificial intelligence software called VUNO Med4-Fundus AI12 to assist in analyzing fundus images for diabetic retinopathy detection in patients with type 2 diabetes mellitus. The study aims to see how effective this AI tool is compared to routine clinical care without AI assistance in identifying patients who need referral for diabetic retinopathy. This is a randomized controlled trial led by VUNO Inc. focusing on improving diabetes eye care through technology. Participants will be randomly assigned to one of two groups. Those in the intervention group will have their fundus images analyzed using the VUNO Med4-Fundus AI12 software, and the AI results will be shared with clinicians during routine care. The control group will receive usual clinical care without the use of AI assistance. The study measures outcomes within 6 months, including true referral rates and other diagnostic performance indicators. During the study, participants will undergo fundus photography as part of their routine care, with or without AI support depending on their group. Researchers will monitor how accurately diabetic retinopathy is detected, the time to diagnosis, referral rates, and adherence to recommendations. Data will be collected over a 6-month period to assess the impact of AI assistance in clinical decision-making for diabetic retinopathy referral. The total study duration extends until August 2027.

Clinical Efficacy of Implementing an AI-SaMD for Funduscopy Analysis in Patients With Diabetes Mellitus