AI cuts diagnosis for eye disease from 23 days to just 30 seconds: NHS set to roll out revolutionary system

**Lead**

A revolutionary artificial intelligence system is set to transform eye care in the United Kingdom, slashing the time needed to diagnose a common sight-threatening condition from more than three weeks to just 30 seconds. The technology, developed through a partnership between Moorfields Eye Hospital NHS Foundation Trust and Google Health, has now been approved for wider NHS deployment, promising to ease pressure on overstretched ophthalmology departments and deliver faster treatment for patients at risk of irreversible vision loss.

**Context**

The condition in question – diabetic retinopathy – affects hundreds of thousands of people in the UK and is a leading cause of preventable blindness among working-age adults. Traditionally, patients referred by their GP or optician must wait an average of 23 days for a specialist to review their retinal scans and confirm a diagnosis. During that window, the disease can progress, and treatment opportunities may be missed. The new AI system analyses optical coherence tomography (OCT) images in real time, flagging urgent cases within seconds and allowing clinicians to prioritise those most in need.

Moorfields Eye Hospital, one of the world’s leading centres for ophthalmic research, has been trialling the AI tool since 2018. Early results, published in peer-reviewed journals, demonstrated that the algorithm could match or even exceed the accuracy of human experts in detecting more than 50 different eye conditions, including diabetic retinopathy, age-related macular degeneration and glaucoma. The latest iteration of the software has been optimised for speed and reliability, with the NHS now planning a phased rollout across hospital trusts in England, Scotland, Wales and Northern Ireland.

The system works by feeding OCT scans – high-resolution 3D images of the retina – into a deep learning model trained on millions of anonymised patient records. Within seconds, it produces a triage recommendation: urgent referral, routine follow-up, or no abnormality detected. This allows ophthalmologists to focus their expertise on complex cases while the AI handles the initial screening burden. According to NHS England, the technology could free up thousands of clinician hours each year and reduce the backlog of patients waiting for eye appointments, which currently stands at over 600,000 across the country.

**Analysis**

The implications of this development extend far beyond convenience. For patients, the difference between a 23-day wait and a 30-second diagnosis can mean the preservation of sight. Diabetic retinopathy, for instance, is highly treatable if caught early, but once the macula is damaged, vision loss is often permanent. By accelerating the diagnostic pathway, the AI system enables earlier intervention with laser therapy or anti-VEGF injections, dramatically improving outcomes. The Royal College of Ophthalmologists has welcomed the move, noting that the technology addresses a critical bottleneck in the care pathway.

From a health system perspective, the cost savings are substantial. Each delayed diagnosis carries a hidden price tag: more advanced disease requires more intensive treatment, longer hospital stays and greater social care support. The AI tool, by contrast, is relatively inexpensive to deploy and can be integrated into existing imaging equipment. NHS Digital estimates that widespread adoption could save the health service tens of millions of pounds annually, while also reducing the number of people who go blind unnecessarily. This aligns with the government’s broader digital health strategy, which aims to embed artificial intelligence across primary and secondary care by 2030.

However, experts caution that technology alone is not a panacea. The AI system must be rigorously monitored for bias, particularly across different ethnicities and age groups, and clinicians must remain the final decision-makers. There are also concerns about data privacy and the need for robust cybersecurity, given the sensitive nature of medical images. Nevertheless, the Moorfields-Google Health collaboration has set a global benchmark for responsible AI deployment in healthcare, and the UK is now poised to become a world leader in AI-assisted ophthalmology.

**Sources**

- Moorfields Eye Hospital NHS Foundation Trust - Google Health (formerly DeepMind) - NHS England - The Royal College of Ophthalmologists - NHS Digital