Majda Hadziahmetovic (Duke University) and Jorge Rocha (iRetina Eye Institute)
Artificial intelligence: a revolution in progress
“Artificial Intelligence (AI) is a tool that will help us be more productive, healthier, smarter, and entertained,” remarked Sam Altman, CEO of OpenAI, the creator of ChatGPT, now the most recognized generative AI. The dawn of a new era is upon us, and AI is here to stay.
The journey of AI began in the summer of 1956 at Dartmouth University, where John McCarthy and colleagues held a historic workshop that coined the revolutionary concept of machines mimicking human learning and decision-making. Three years later, Arthur Samuel advanced this vision with the birth of Machine Learning (ML)—algorithms enabling computers to learn and adapt without explicit programming. ML reached new heights with Deep Learning (DL), a method leveraging multi-layered neural networks that mimic the human brain. Techniques such as Convolutional Neural Networks (CNNs) have since driven breakthroughs in image recognition, including life-changing applications in medicine.
The ability of AI to emulate human learning and continuously enhance itself is both inspiring and unsettling. AI has evolved into an indispensable partner in human progress, enhancing productivity in different fields, especially in medicine.
The role of AI in modern medicine
AI has proven transformative in healthcare, offering cost-effective solutions for diagnostics, disease progression and response to treatment prediction, offering personalized treatment. AI has made remarkable strides in addressing diabetic retinopathy (DR), the leading cause of vision loss among working-age adults. With global diabetes cases projected to surpass 600 million by 2040, AI presents an opportunity to improve access and bridge the gap in DR screening, a critical need given that over 50% of cases are diagnosed too late for optimal intervention.
Traditionally, DR diagnosis relies on a “human in the loop” and “store and forward” approach, where experts analyze retinal images for signs of pathology. However, autonomous AI-powered devices now provide comparable accuracy, offering a scalable solution for early detection.
FDA approved AI solutions for DR
In 2018, the FDA approved the first autonomous AI device for DR screening, IDx-DR. Since then, other algorithms like EyeArt and AEYE-DS have achieved regulatory approval, demonstrating diagnostic sensitivity and specificity comparable to expert ophthalmologists. IDx-DR (LumineticsCore) demonstrated sensitivity of 87.2% and specificity of 90.2% in clinical trials. In recent studies, EyeArt performed with a sensitivity of 95.5% and a specificity of 85.0%. In addition, AEYE-DS also received FDA clearance in 2022 for its robust performance metrics.
These technologies, already getting integrated into U.S. and European healthcare systems, represent the next frontier: large-scale validation in high-volume hospital settings to demonstrate cost and clinical effectiveness.
The future of AI in ophthalmology
AI-powered DR screening offers immense potential to reduce blindness worldwide. Yet, it sparks mixed reactions that prevent its wider adoption. Regardless, the evidence is clear – AI-supported retina screening has the potential to transform (diabetic) retinal care and, by proxy, public health. It can help efficacy and value-based care, save money, and, last but not least, improve clinical outcomes.
References:
- McCarthy J, Minsky ML, Rochester N, Shannon CE, A proposal for the Dartmouth summer research project on artificial intelligence. AI Magazine. 2006;27(4)
- Samuel AL. Some studies in machine learning using the game for checkers. IBM J res Dev, 1959 Jul;3(3):210-29.
- Wu CT, Lin TY, Lin CJ, Hwang DK. The future applications of artificial intelligence and telemedicine in the retina: A perspective. Taiwan J Opthalmol 2023; 13:133-41.
- Sheng B, Cheng X, Li T, Ma T, Yang Y, Bi L and Zhang X (2022). AN overview of artificial intelligence in diabetic retinopathy and other ocular diseases. Front. Public Health 10:971943.
- Rajest AE, Davidson OQ, Lee CS, Lee AY. Artificial intelligence and diabetic retinopathy: AI framework, prospective studies, head-to-head validation and cost-effectiveness. Diabetes Care 2023;46:1728-1739.
- van der Heijden AA, Abramoff MD, Verbraak F, van Hecke MV, Liem A, Nijpels G. Validation of automated screening for referable diabetic retinopathy with the IDx-DR device in the Hoorn Diabetes care System. Acta Ophthalmol.(2018) 96:63-8.
