Artificial intelligence to predict and prevent disease

In summary

Professors Chris Owen and Alicja Rudnicka are leading the development of AI-powered technology to analyse images from eye scans, in a bid to detect a range of conditions before people show symptoms.

They have already proven that their AI tools can identify differences in the shape and size of blood vessels at the back of the eye, which correspond to cognitive status, high blood pressure and heart disease.

One major focus is to evaluate AI approaches to detect risk of diabetic eye disease - a major complication of diabetes. This occurs when the delicate network of tiny blood vessels at the back of the eye becomes damaged. It can lead to blindness if undetected and untreated.

The number of people living with diabetes is large and increasing, putting a major burden on the NHS in terms of cost and resource.

Everyone with diabetes is referred for an eye check often every one or two years.

Numerous images are taken at each visit, generating millions of images each year that are manually analysed by specialists.

But AI is proving it can detect the presence of sight-threatening diabetic eye disease risk from these images as well as humans can, and very rapidly, which could help alleviate pressure on the diabetic eye screening service.

What did we explore and how?

An AI algorithm developed by the team in the School of Health and Medical Sciences works by extracting tens of thousands of vessel measurements from retinal images, including vessel width, area, and curvature, many of which are invisible to the human eye. By analysing these data, the system identifies early signs of disease.

Working in collaboration with partners including Kingston University, Moorfields Eye Hospital, and Homerton Healthcare NHS Foundation Trust, the algorithms have then been tested on thousands of patients, proving that they can help detect signs of dementia and Alzheimer’s, diabetic eye disease and heart disease.

Benefits and influence of this research

The development of this AI technology has huge scope for it to be seamlessly embedded into the daily workflows of opticians on the high-street and in eye clinics.

It aims to be a major aid to healthcare professionals by speeding up the process of analysing eye scans to just seconds so that time and expertise can be put to better use.

By deploying the AI, it is expected to save the NHS millions of pounds every year, whilst enabling patients to get a diagnosis, support and access to vital treatment much sooner.

This innovation could help contribute to the government’s 10 Year Health Plan to shift the focus of the NHS out of hospitals and into the community in routine healthcare settings.

AI analysis of eye images could transform healthcare by enabling early disease detection, when treatments are most effective. With retinal imaging already widely available in clinics and even on the high street, this technology has the potential to be a game changer.