An AI model called RETFound, developed by a team at Moorfields Eye Hospital and UCL Institute of Ophthalmology, has the potential to predict a person’s risk of having a heart attack by analyzing their retinal scans. The model is trained using vast amounts of data from the NHS and can also detect other sight-threatening diseases such as diabetic eye disease, glaucoma, Parkinson’s disease, and stroke. The developers believe that this AI model could have significant public health implications and contribute to the UK’s ambition of becoming an AI “superpower.”
Professor Pearse Keane, an expert in artificial medical intelligence at UCL, explained that the combination of big data, advanced imaging, and AI could “supercharge” the ability to detect signs of systemic diseases by examining the eyes. If successful, this technology could revolutionize eye tests by providing insight into a person’s risk of having a heart attack in the next five years.
AI models like RETFound have gained attention for their transformative potential. They analyze data to identify patterns and make predictions. RETFound specifically uses retinal scans from millions of NHS patients to make predictions about various conditions.
The RETFound model has been made open source, allowing researchers around the world to use the algorithm and adapt it for different populations. This enhances the potential for its application in diverse healthcare settings and ensures that the AI system is reliable and fair across different ethnic groups.
The development of RETFound aligns with the UK government’s National Artificial Intelligence Strategy, which aims to leverage technology to address public health challenges. Professor Keane emphasized that RETFound showcases the collaboration between world-leading NHS data and top computer science expertise from UK universities, positioning the UK as a potential AI superpower in the healthcare sector.
In addition to predicting heart attack risk, the researchers also found that AI-powered retinal scans could be used to detect Parkinson’s disease. By analyzing retinal markers, the AI model successfully identified patients with Parkinson’s, suggesting the potential for this technology to be used as a pre-screening tool for early detection and intervention.
Sources:
– Original article by Storm Newton, Health Reporter on PA Media
– More information on RETFound AI model and the collaboration between Moorfields Eye Hospital and UCL Institute of Ophthalmology
