Type 1 diabetes (T1D) brings many challenges, and one of the most critical is protecting your vision. Diabetic Eye Disease (DED) is a leading cause of blindness, and early detection is crucial, especially for young people with T1D. Many in underserved communities do not have equitable access to essential screenings, widening the care gap and increasing the risk of severe eye problems.

In a groundbreaking study funded by Diabetes Research Connection (DRC), Dr. Risa Wolf at Johns Hopkins University is leveraging the power of autonomous artificial intelligence (AI) to transform early detection of DED in T1D youth. Dr. Wolf’s research focuses on using AI to boost screening rates in these communities, offering a promising solution to improve outcomes for T1D youth at risk of DED. Her study reveals that AI significantly increases diabetic eye exam completion rates, effectively closing the care gap in communities with fewer resources. This breakthrough enhances early detection and improves follow-up care, creating a more accessible and inclusive approach to DED management.

 For more insights on maintaining eye health with T1D, don’t miss our other blog this month, “Protecting Your Vision: Essential Eye Care for T1D.” This blog delves into crucial strategies and practical tips for preserving vision and preventing diabetes-related complications. It covers everything from regular eye exams and managing blood sugar levels to recognizing early signs of diabetic retinopathy and other vision issues.

Q&A with Dr. Risa Wolf

Understanding the innovative work of Dr. Risa Wolf can provide invaluable insights into the future of DED management for T1D. To delve deeper into her research and its implications, Dr. Wolf answers questions that explore her background, the study’s findings, and the potential impact on T1D care.

Could you share your journey and what inspired you to focus on pediatric endocrinology, especially diabetic eye disease?

When I was a resident at the Children’s Hospital of Philadelphia, I was fortunate to have an outstanding supervising physician on the inpatient endocrine service whose love and excitement for the complexity of endocrine pathways and associated disorders inspired me to pursue a career in pediatric endocrinology. After completing my fellowship in pediatric endocrinology at Johns Hopkins, I relocated to South Florida to start a new pediatric endocrine practice for a large hospital system. There, I quickly recognized the need for comprehensive and multidisciplinary diabetes care, leading to the establishment of a new diabetes center at Broward Health that included an endocrine provider, diabetes nurse, nutritionist, and psychologist. To provide comprehensive care in one place, I began incorporating diabetic eye disease screening using a handheld fundus camera. Subsequently, I returned to Johns Hopkins Hospital for a faculty position and continued my efforts to integrate diabetic eye disease screening into routine diabetes care.

What have been some of the most rewarding moments in your research career, particularly at Johns Hopkins University?

There are many rewarding components to being a physician-investigator, including the opportunity to do research that impacts clinical care and outcomes, and mentoring students, residents, and fellows through successful completion of a research project. One of the most exciting moments of my research career was being awarded my first National Institutes of Health (NIH) grant. This was particularly rewarding because I started out with seed funding from Diabetes Research Connection and a small innovation grant from Johns Hopkins to conduct our first autonomous AI study, and the results led to this larger NIH grant.

How does the autonomous AI system work in detecting early signs of DED in young patients with T1D?

The autonomous AI system (Digital Diagnostics, formerly IDx-DR) that we use in pediatrics is a lesion-based algorithm, meaning that it looks for specific markers of diabetic retinopathy, such as microaneurysms and hemorrhages, on the images.  If it detects an abnormality consistent with diabetic retinopathy, it results in an abnormal diabetic eye exam, with the recommendation for eye care follow-up.

What significant changes have you observed in the early detection rates of DED since implementing AI in your study?

Early detection of diabetic retinopathy leads to treatment that can prevent worsening of eye disease and preserve visual health.  Implementation of AI diabetic eye exams at the point of care increases screening rates for all; and more screening can lead to identification of unknown diabetic eye disease, particularly for individuals without prior eye screening or those who have had a lapse in completing regular diabetic eye exams.

Can you elaborate on how AI technology helps to bridge the care gap in under-resourced communities?

There are well-known disparities in diabetic retinopathy screening, with minority and underserved individuals less likely to complete the diabetic eye exam.  From our early studies and clinical experience, we learned that minority youth in our clinic were less likely to have completed a diabetic eye exam, yet more likely to have diabetic retinopathy.  Our first use of the autonomous AI system in our clinic improved screening rates for all patients, and in the recent AI for Childrens diabetiC Eye examS Study (ACCESS2) trial, we demonstrated that implementation of AI mitigates disparities in screening completion for minority youth, promoting health equity for under-resourced communities.

What were the diabetic eye exam completion rates before and after the introduction of AI in your study?

Prior to implementing the point-of-care AI diabetic eye exams in our clinic, our screening rates were around 50%.  We explored barriers to diabetic eye exam completion in our population and found the most cited barriers were 1) not recalling being recommended to obtain a diabetic eye exam, 2) difficulty finding time for an additional appointment, and 3) transportation issues. After implementing AI, our screening rates improved to 95%.

How has AI influenced the follow-up care process for youths diagnosed with early signs of DED?

The AI system provides a diagnosis right away, once the brief eye exam is complete, and this has been shown by our team and others to improve the likelihood that patients’ follow up with an eye care provider when an abnormality is detected.

With AI playing a critical role in diagnosis, how do you ensure the accuracy and reliability of the AI system in clinical settings?

There are several AI systems to diagnose diabetic retinopathy that have FDA approval and have met specified measures for accuracy and reliability.  These systems have also been tested in real-world clinical settings with similar diagnostic accuracy.  As the first diabetes center to use an AI system for diagnosing diabetic retinopathy in the pediatric population, we took additional steps to validate the diagnostic accuracy of the algorithm in youth.

Based on your research findings, how do you see AI evolving in managing diabetes and its complications over the next decade?

Diabetes care is one of the few fields that effectively utilizes AI in various innovative ways, including hybrid-closed loop insulin systems, smart pens, and autonomous AI for diabetic retinopathy. These AI systems enhance glycemic control for users and expanding their accessibility across all races and ethnicities can help promote health equity. I believe this is just the beginning, and we will see even greater integration of these systems in the future.

What advice would you give young researchers interested in leveraging technology to address healthcare disparities?

Be open to learning about and embracing new technologies, and consider their potential impact on patients. It’s crucial to evaluate the feasibility and acceptability of these technologies, while also ensuring safety, privacy, and effectiveness for the specific populations you serve. Thoughtful implementation can help address healthcare disparities and improve outcomes for diverse patient groups.

Dr. Wolf’s answers offer a fascinating glimpse into how AI can revolutionize DED detection and management. Her work highlights the importance of cutting-edge research in improving health outcomes for T1D patients, especially in underserved communities. By supporting initiatives like those funded by DRC, we can contribute to significant advancements in T1D care, ensuring better health and quality of life for all.

The Future of Diabetic Eye Disease Management

Dr. Risa Wolf’s pioneering research is a beacon of hope for those managing T1D. By harnessing the power of AI, her study significantly enhances the early detection of DED. This breakthrough improves screening rates and ensures better follow-up care, offering a more inclusive approach to managing this serious eye condition.

Supporting research through DRC is crucial for continuing advancements in T1D care. Your donations can help fund innovative studies like Dr. Wolf’s, driving progress in early detection and treatment methods that can transform lives.

For more tips on maintaining eye health with T1D, read our other blog this month, “Protecting Your Vision: Essential Eye Care for T1D.” By staying informed and proactive, we can collectively work towards a future where DED no longer threatens the vision of those living with T1D.