Por Renan Filho · 2026-10-14
**Lead**
A revolutionary artificial intelligence (AI) tool has cut the diagnosis time for a leading cause of blindness from 23 days to just 30 seconds, according to findings published in a peer‑reviewed medical journal. The development, which has been hailed as a potential game‑changer for ophthalmology, is now being closely examined by NHS trusts and the Royal College of Ophthalmologists as a way to tackle mounting waiting lists and reduce preventable sight loss across the United Kingdom.
The AI system, developed by an international research consortium, analyses retinal scans with a speed and accuracy that matches – and in some cases exceeds – that of senior clinicians. In a clinical validation study involving thousands of patients, the technology correctly identified signs of age‑related macular degeneration and diabetic retinopathy in under a minute, compared with the current average wait of more than three weeks for a specialist review. The results have been submitted to the National Institute for Health and Care Excellence (NICE) for evaluation, and early discussions are under way with several NHS hospital trusts about pilot programmes.
**Context**
Eye disease represents a growing burden on the UK’s healthcare system. According to the Royal National Institute of Blind People, more than two million people in the UK live with sight loss, and that figure is projected to rise sharply as the population ages. Diabetic retinopathy alone affects around one in three people with diabetes, and age‑related macular degeneration is the leading cause of blindness in those over 50. Early diagnosis is critical: delays of even a few weeks can mean the difference between preserving vision and irreversible damage.
Currently, patients referred for suspected retinal disease typically undergo an optical coherence tomography (OCT) scan at a high‑street optician or hospital clinic. The images are then sent to a hospital ophthalmology department, where they join a queue for review by a consultant. With the NHS facing a shortage of more than 500 ophthalmologists, waiting times have stretched to an average of 23 days – and in some regions, up to six weeks. During that period, a patient’s condition can deteriorate significantly.
The new AI system, trained on a dataset of more than 100,000 anonymised retinal scans from multiple countries, can analyse an OCT image in real time. It flags abnormalities, grades the severity of disease, and generates a report for the clinician – all within 30 seconds. The technology does not replace the doctor but acts as a triage tool, prioritising urgent cases and allowing routine scans to be handled more efficiently.
**Analise**
The implications for the UK are profound. The NHS performs over 10 million OCT scans each year, and the backlog of unread images has become a major bottleneck. By integrating AI into the diagnostic pathway, hospitals could reduce the time from scan to treatment from weeks to hours. This would not only improve patient outcomes but also free up consultant time for complex cases and surgical procedures.
Moreover, the economic case is compelling. Sight loss costs the UK economy an estimated £28 billion annually in direct healthcare costs, social care, and lost productivity. Every week of delayed diagnosis increases the likelihood of irreversible damage, leading to higher long‑term costs. If the AI system can prevent even a fraction of these delays, the savings could run into hundreds of millions of pounds. The technology is also relatively inexpensive to deploy, as it runs on standard hospital computers and does not require expensive hardware upgrades.
However, adoption is not without challenges. The AI must be rigorously validated on UK‑specific patient populations, which may have different disease prevalence and ethnic variations. Data protection and patient consent are also critical concerns, as the system relies on cloud‑based processing in some configurations. The Royal College of Ophthalmologists has called for a cautious, evidence‑based rollout, emphasising that AI should augment – not replace – clinical judgement. Pilot programmes are expected to begin within the next six months at selected NHS trusts, with a wider rollout possible by 2028 if the technology meets NICE’s standards.
**Fontes**
- Royal National Institute of Blind People (RNIB) – sight loss statistics and economic impact data. - Royal College of Ophthalmologists – workforce shortages and AI adoption guidance. - National Institute for Health and Care Excellence (NICE) – evaluation process for medical technologies. - Published study in The Lancet Digital Health (2026) – validation of AI system for retinal disease diagnosis. - NHS England – ophthalmology waiting times and referral data.