Pediatric endocrinologist Risa Wolf discusses the latest technologies for improving the care and outcomes for patients with diabetes. The latest artificial intelligence technology for pediatric diabetic retinopathy screenings is used at Johns Hopkins. By screening patients early, doctors are able to help them avoid complications from this disease, which include vision loss and blindness.
Hi, my name is Risa Wolf and I'm an Associate professor of pediatrics at the Johns Hopkins Children's Center. My research focuses on using innovative technologies to improve the care and outcomes for Children with diabetes. One of the complications of diabetes is diabetic retinopathy that can lead to vision loss and blindness. Screening for diabetic retinopathy with a diabetic eye exam can prevent this but only about 50% of Children undergo the recommended eye exams and Children from underserved communities are even less likely to undergo these recommended screenings but also have an increased risk of developing diabetic retinopathy. While the traditional diabetic eye exam includes going to an eye care provider for a dilated eye exam which can take several hours. The goal of our work is to increase screening and early detection of diabetic retinopathy using new and innovative autonomous Artificial Intelligence or AI technology, which can be done at the point of care using a Nondemocratic Funders camera that takes four pictures of the eye without dilation and then the autonomous AI system provides an immediate result. So the entire diabetic eye exam is complete before leaving the diabetes care visit. We were the first pediatric diabetes center to use this AI technology and showed it was safe for use in Children in a pilot study of over 300 patients. We also found that doing point of care screening helped us to increase our screening rates from 49 to 95%. We called this the C study safety efficacy and equity of diabetic retinopathy screening and youth. In this study, we also found that minority youth were less likely to have had a prior diabetic eye exam, yet more likely to have diabetic retinopathy. So in further studies called access AI for children's diabetic eye exams, we are conducting a larger multi site study within Johns Hopkins medicine to determine if implementing point of care diabetic retinopathy screening is more effective and can mitigate racial ethnic disparities in access to the diabetic eye exam. This autonomous AI system was the first diagnostic AI system to receive FDA approval as an early user of the system. Our research team has expanded our work to include ethical considerations of autonomous AI and clinical trials, cost savings analysis and long term visual outcomes And considering the patient and physician perspective on integrating these novel technologies into practice. While the prevalence of diabetic retinopathy is lower in childhood than in adults. Up to 5% of youth with type one, diabetes can have diabetic retinopathy and this can increase in adulthood. Recent data and youth onset type two diabetes shows that within the first few years of having type two diabetes. The prevalence is approximately 10 to 15% but can be as high as 50% after having diabetes for 12 to 13 years. With these astonishingly high rates, we are hoping that we can increase screening and early detection of diabetic retinopathy to prevent blindness in this young age group. We are very grateful to the Children's Center for our initial funding to support this research as well to the diabetes research connection in the National Eye Institute. Thank you so much for listening today.