In a development that could significantly reshape eye care, researchers have developed an artificial intelligence (AI)-based system, OCTCube-M, capable of analysing complex retinal scans with greater speed and accuracy.
The experimental technology, potentially helping ophthalmologists detect vision-threatening diseases at the earliest, has been developed by researchers from Washington University School of Medicine in St. Louis, in collaboration with scientists from the University of Washington and biotechnology company Genentech.
Published in Nature Biomedical Engineering, the findings suggest that the system can interpret detailed three-dimensional retinal images more effectively than existing AI models, thus offering new possibilities for both patient care and medical research.
The advance comes at a time when retinal diseases such as age-related macular degeneration (AMD), diabetic retinopathy, and glaucoma are emerging as major causes of visual impairment globally, including in India, where rising diabetes rates and an ageing population are contributing to a growing burden of eye disease.
“Today's eye scans provide physicians an unprecedented, highly detailed view of the inside of the eye, revealing structures and subtle changes that would otherwise go undetected,” said Dr. Aaron Lee, Arthur W. Stickle Distinguished Professor of Ophthalmology and Visual Sciences at Washington University School of Medicine.
“But we still lack the tools to help physicians process the volume of generated images. Our AI system has the potential to empower physicians to make faster diagnoses, tailor treatment more precisely, and design clinical trials that bring new therapies to patients faster,” he said.
The new system was trained using more than 26,000 three-dimensional OCT scans, comprising over 1.6 million individual retinal images. Unlike earlier AI models that relied largely on two-dimensional images, OCTCube-M analyses the retina in three dimensions, enabling a more comprehensive understanding of disease patterns.
Researchers found that the AI system improved detection of six out of eight major retinal diseases, identifying approximately 43 to 60 additional cases per 1,000 patients compared with earlier models.
Such improvements may appear modest numerically, but experts note that in large populations, these gains could translate into thousands of patients receiving earlier diagnosis and treatment, potentially preventing irreversible vision loss.
By combining OCT scans with additional imaging techniques such as infrared retinal imaging and fundus autofluorescence, researchers found that the AI model predicted disease progression nearly 50% more accurately than current state-of-the-art systems.
This capability could have far-reaching implications for drug development.
“By better predicting how fast disease will worsen, we can run smaller, more efficient studies,” Dr. Lee explained. “That could lower costs, shorten the time it takes to test new therapies, reduce the number of people exposed to treatments that don't work, and help effective drugs reach patients sooner.”
Beyond eye diseases, the researchers reported another intriguing possibility. Because the retina contains tiny blood vessels that reflect changes occurring elsewhere in the body, the AI system was also able to identify patterns associated with broader health risks, including cardiovascular disease, stroke, and kidney failure.
“The model has the potential to turn a simple eye exam into a powerful tool for helping to detect illness beyond the eye,” Dr. Lee said.
The findings are particularly relevant for India, where diabetic retinopathy and age-related retinal disorders are becoming increasingly common. Limited access to retinal specialists in many districts often results in delayed diagnosis, by which time significant vision loss may already have occurred.
AI-assisted screening tools could help bridge this gap by supporting ophthalmologists and enabling faster interpretation of scans in busy clinics and telemedicine networks.
The study, however, cautions that the technology remains in the research phase and is not yet ready for routine clinical use. Larger datasets involving more patients, diseases, and imaging modalities will be needed before regulatory approval and widespread adoption.
India has been witnessing a growing integration of artificial intelligence in ophthalmology. At institutions such as the All India Institute of Medical Sciences (AIIMS), Delhi, researchers and clinicians have been evaluating AI-assisted tools for screening diabetic retinopathy, glaucoma, and other retinal disorders using fundus photographs and retinal scans.
