An artificial intelligence system can now identify advanced age-related macular degeneration (AMD) with 90% accuracy using routine retinal photographs taken in primary care clinics, according to a new study. The breakthrough could help catch this sight-threatening eye disease earlier, before patients experience significant vision loss.

What Is This New AI Screening Tool?

Researchers tested an AI system called iPredict-AMD on 696 adults over age 50 across six clinics in New York City. The system analyzed standard retinal images captured without pupil dilation, a step that can make clinic visits longer and cause temporary light sensitivity. The AI correctly identified 102 out of 113 cases of intermediate or advanced AMD, meaning it caught the disease when it was most likely to benefit from treatment.

What makes this particularly exciting is the speed and accessibility. The system returned results within one minute and didn't require the dilating drops that many eye exams involve. The AI also flagged other serious retinal conditions it wasn't specifically trained to detect, including diabetic retinopathy and macular puckering, suggesting it could catch multiple vision problems in a single screening.

Why Does Early Detection of Macular Degeneration Matter?

Age-related macular degeneration affects the macula, the small area at the center of the retina responsible for sharp, detailed vision needed for reading, recognizing faces, and driving. When AMD progresses from early to intermediate or advanced stages, vision loss can accelerate significantly. Early detection allows doctors to start treatments that can slow or prevent further damage, potentially preserving sight for years longer.

The study found that the AI system had a negative predictive value of 97.79%, meaning when it said someone didn't have advanced AMD, that assessment was correct nearly 98% of the time. This high accuracy matters because it means fewer people would need unnecessary specialist referrals, while those with disease wouldn't slip through the cracks.

How Does This Change Eye Care in Primary Care Settings?

Traditionally, screening for AMD happens in ophthalmology offices, which means patients need a referral and often face long wait times. By bringing this screening into primary care clinics, where many people already visit for routine checkups, the technology could catch AMD earlier in its course. The AI system works with standard equipment that many primary care offices already have or could easily acquire.

However, researchers identified real barriers to widespread adoption. Rural and low-resource clinics may lack the imaging devices needed, sufficient internet bandwidth for cloud-based processing, or staff trained to operate the system. These limitations mean the technology's benefits may not reach everyone equally, at least in the near term.

How to Protect Your Vision From Macular Degeneration

• Get Regular Eye Exams: Adults over 50 should have comprehensive eye exams at least every one to two years, or more frequently if recommended by their eye doctor. Early-stage AMD often has no symptoms, so screening is essential.
• Know Your Risk Factors: Age, family history of AMD, smoking, high blood pressure, and obesity all increase your risk. If you have these factors, discuss more frequent screening with your doctor.
• Ask About Advanced Imaging: If your primary care doctor offers retinal imaging or AI screening, take advantage of it. These tools can catch disease before you notice vision changes.
• Monitor Your Vision at Home: Use an Amsler grid (a simple checklist pattern) to test your central vision weekly. Report any new distortion, blurriness, or dark spots to your eye doctor immediately.

What Other Eye Conditions Can This Technology Detect?

Beyond AMD, the AI system unexpectedly identified cases of diabetic retinopathy, epiretinal membrane, high myopia with retinal concerns, and macular pucker. This suggests the technology could serve as a broader screening tool for multiple retinal diseases, not just AMD. People with diabetes, in particular, could benefit from this multi-condition detection capability.

The broader lesson is that modern imaging technology, combined with artificial intelligence, is making it possible to screen for serious eye diseases in settings where patients already receive care. This shift could democratize early detection and help preserve vision for millions of people who might otherwise not see an eye specialist until damage is already done.

If you're over 50 or have risk factors for AMD, ask your primary care doctor whether they offer retinal imaging or AI-assisted screening. Catching this disease early, before symptoms appear, remains the best way to protect your sight.