St George's, University of London and Moorfields Eye Hospital NHS Foundation Trust are leading a half a million pound project on artificial intelligence (AI), funded by NHSX and the Health Foundation, and enabled by the National Institute for Health Research (NIHR). The research project will use AI technology to analyze retinal images from diabetic patients to detect sight-threatening diabetic eye disease, more efficiently and quicker than human specialists. Crucially, the project will develop safeguarding systems to ensure it works for all and that AI performance does not vary across population sub-groups, such as ethnicity or gender. The project will also provide evidence to support the commissioning and deployment of the first potential widespread use of AI within the NHS.
Researchers at St George's and Moorfields have previously shown that automated retinal image analysis systems using AI can indeed provide a safe, cost-effective alternative. This could provide major benefits for the NHS Diabetic Eye Screening Programme (DESP), which generates millions of retinal images every year for early detection of diabetic eye disease - the leading causes of blindness in the working age population.Considering the cost and number of screening episodes, the savings could extend to more than £10 million every year in England alone.
Concerns have been raised over AI image recognition technologies, such as facial recognition varying in performance with gender and ethnicity. Researchers at St George's and Moorfields will create a database of retinal images from different ethnic groups, genders and ages to ensure continued safety and confirm that anyone affected by diabetes in the UK will benefit. They will convene a group from North East London DESP including diabetic patients, a screening specialist, a consultant endocrinologist and a public health physician, informing the best way to use these systems within the NHS DESP.
It's crucial that the first wide-scale deployment of AI in the NHS is safe and performs to a high level across the board. This project will develop the essential safety tools necessary, and monitoring systems to check the performance of AI after deployment to ensure trustworthy AI for the benefit of all patients."
Professor Adnan Tufail, Consultant Ophthalmologist at the NIHR Biomedical Research Centre at Moorfields Eye Hospital and UCL Institute of Ophthalmology
"Importantly, this project will also be independent of any commercial interests," said Professor Alicja Rudnicka, Professor of Statistical Epidemiology at St George's, University of London. "We will also evaluate the perceptions, acceptability and expectations of health care professionals and people with diabetes in relation to AI technology implementation within the North East London DESP. Moreover, the methodology and standards we develop will be transferable to other healthcare domains to build trust in AI technology in healthcare moving forward."
This project will commence in 2021 in partnership with The Homerton University Hospital, Kingston University, and University of Washington, USA.