NHS Unveils AI Tool Predicting Diabetic Retinopathy

NHS researchers at King’s College London have developed an advanced AI model to predict the risk of diabetic retinopathy up to three years in advance. Leveraging data from the NHS Diabetic Eye Screening Program (DESP) and over a million retinal images, this innovative tool aims to identify individuals at high risk, potentially saving the NHS millions and enhancing patient care.

The NHS Diabetic Eye Screening Program annually screens around 3.2 million diabetes patients in the UK, costing approximately £85 million. This preventive measure is crucial given diabetic retinopathy’s potential to cause vision loss.

The new AI model, derived from extensive image data provided by the Southeast London DESP and validated with 70,000 images from the INSIGHT health data research hub, identifies high-risk patients with remarkable accuracy. INSIGHT, an NHS-led organisation, integrates retinal images with clinical data from Moorfields Eye Hospital and University Hospitals Birmingham NHS Foundation Trust, covering contributions from over 200,000 patients. By predicting patients’ risk levels one to three years ahead, this AI model promises to optimize screening schedules, focusing resources on those most in need and reducing unnecessary checks for low-risk individuals.

The new AI-powered model marks a significant advancement in diabetic retinopathy management, with promising implications for patient outcomes and NHS resource allocation. Future clinical trials involving over 50,000 diabetes patients will determine its safety, efficacy, and cost-effectiveness, potentially revolutionizing diabetic eye care across the UK.