DeepMind emphasises role of AI in detecting eye disease at HIMSS Impact18

Dr Dominic King will continue to lead the Streams team.
By Tonya Stewart
11:29 am

One of the most popular keynotes at HIMSS’ annual Impact event in Potsdam, just outside Berlin, Germany last month was Dominic King’s AI-enabled healthcare: potential and challenges, which emphasised AI’s wide-ranging potential in revolutionising everything from imaging to oncology – underlining in particular its revolutionary potential for detecting eye conditions early. 

The AI research firm’s groundbreaking research with Moorfields Eye Hospital, London, DeepMind Health's Clinical Lead said, had been a highlight. In an interview after the event, King told MobiHealthNews:

“In August this year, Nature Medicine published findings from our initial research with the hospital, which set out to build an AI system that could detect conditions of eye disease.

“Hospital eye services in the UK see close to 10 million patients annually – a figure that is only second to orthopaedics and is growing year on year. The vast majority of cases of sight loss caused by retinal diseases, such as age-related macular degeneration, glaucoma and diabetic retinopathy, are entirely curable if they are caught early enough.

“Currently, eye care professionals use optical coherence tomography (OCT) scans to diagnose eye conditions, but these 3D images are hard to read and need expert analysis to interpret. This process all takes time and there aren’t enough qualified doctors to deal with the sheer volume of cases. 

“Our research model can automatically detect over 50 types of eye disease in seconds, matching the level of experts with over 20 years’ experience, and can help clinicians prioritise patients in most need of urgent care,” King said. 

“Crucially, the model didn’t miss a single urgent case in the test set of 1000 cases which were selected by stratification and tested retrospectively.”