A recent Australian trial has demonstrated that artificial intelligence (AI) can effectively detect diabetic eye disease, potentially transforming routine clinical care for individuals living with diabetes. Conducted by a team of researchers including Associate Professor Lisa Zhuoting Zhu and Sanil Joseph from the Centre for Eye Research Australia and University of Melbourne, alongside Professor Mingguang He from the Hong Kong Polytechnic University, the study highlights the promising role of AI in increasing access to sight-saving eye screenings.
Globally, over 529 million people are affected by diabetes, which significantly raises the risk of vision loss and blindness due to diabetic eye disease. Early intervention can prevent blindness in up to 90 percent of cases; however, ensuring widespread access to necessary eye scans remains a significant challenge for health systems worldwide.
The findings from a two-year trial, published in the British Journal of Ophthalmology, indicate that AI could play a crucial role in overcoming these barriers. The trial involved more than 860 participants with diabetes who were recruited from general practice and endocrinology clinics in Melbourne and an Aboriginal Health Service in Western Australia between August 2021 and June 2023.
Participants utilized an automated portable retinal camera, powered by a sophisticated AI algorithm trained on over 200,000 retinal images graded by 21 ophthalmologists. This technology allowed individuals to take photographs of their own eyes while waiting for their medical appointments. They received a report with a QR code linking to their scan results, which facilitated referrals to eye care specialists for those exhibiting signs of diabetic eye disease.
To assess accuracy, results from the AI scans were compared to the traditional gold standard of human grading. Additionally, participants and healthcare professionals participated in a satisfaction survey, contributing valuable feedback on the trial’s implementation. The Melbourne study stands out as one of the first to evaluate AI in real-world clinical settings for diabetic eye disease.
The results revealed several opportunities for improvement. Dr. Zhu noted the significant benefits of AI scans in rural and remote areas, where there is often a shortage of trained eye care specialists. “AI scans could be a great benefit in rural and remote areas,” she stated. “It is also a cost-saving measure for the health system, as it enables early screening to occur without requiring an eye care specialist for every patient.”
Moreover, Sanil Joseph emphasized the convenience of AI-powered eye scans for patients. “People with diabetes often have many medical appointments and prioritize appointments with other specialists over eye care. The AI scan enables them to combine their eye test with other medical visits,” he explained.
This project received funding from the Medical Research Future Fund, and the Centre for Eye Research Australia benefits from Operational Infrastructure Support from the Victorian State Government. The ongoing integration of AI in healthcare signifies a promising step toward improving accessibility and efficiency in eye disease detection.
As health systems continue to adapt to the increasing number of individuals living with diabetes, the potential for AI to reshape the landscape of diabetic eye disease screening remains significant. The findings of this trial may pave the way for broader implementation of AI technologies, ultimately enhancing patient care and reducing the risk of vision loss for millions worldwide.
