Using Artificial Intelligence to Detect Eye Disease
Researchers at the Moorfields Eye Hospital NHS Foundation Trust, DeepMind Health and University College London (UCL) have developed an artificial intelligence (AI) system that can recommend the correct referral decision for over 50 eye diseases as accurately as world-leading experts, it is claimed.
The breakthrough research, published online by Nature Medicine, describes how machine learning technology has been successfully trained on thousands of historic de-personalised eye scans to identify features of eye disease and recommend how patients should be referred for care.
It is hoped that the technology could one day transform the way professionals carry out eye tests, allowing them to spot conditions earlier and prioritise patients with the most serious eye diseases before irreversible damage sets in. More than 285 million people worldwide live with some form of sight loss. Eye diseases remain one of the biggest causes of sight loss, and many can be prevented with early detection and treatment.
The research has been led by UCD Medicine alumnus, Dr Pearse Keane who is a consultant ophthalmologist at Moorfields Eye Hospital NHS Foundation Trust and NIHR Clinician Scientist at the UCL Institute of Ophthalmology.
“The number of eye scans we’re performing is growing at a pace much faster than human experts are able to interpret them. There is a risk that this may cause delays in the diagnosis and treatment of sight-threatening diseases, which can be devastating for patients.
“The AI technology we’re developing is designed to prioritise patients who need to be seen and treated urgently by a doctor or eye care professional. If we can diagnose and treat eye conditions early, it gives us the best chance of saving people’s sight. With further research it could lead to greater consistency and quality of care for patients with eye problems in the future.”
The study, which was launched in 2016, investigated whether AI technology could help improve the care of patients with sight-threatening diseases, such as age-related macular degeneration and diabetic eye disease. Using two types of neural network – mathematical systems for identifying patterns in images or data – the AI system quickly learnt to identify ten features of eye disease from highly complex optical coherence tomography (OCT) scans. The system was then able to recommend a referral decision based on the most urgent conditions detected. To establish whether the AI system was making correct referrals, clinicians also viewed the same OCT scans and made their own referral decisions. The study concluded that AI was able to make the right referral recommendation more than 94% of the time, matching the performance of expert clinicians.
The AI has been developed with two unique features which maximise its potential use in eye care. Firstly, the system can provide information that helps explain to eye care professionals how it arrives at its recommendations. This information includes visuals of the features of eye disease it has identified on the OCT scan and the level of confidence the system has in its recommendations, in the form of a percentage. This functionality is crucial in helping clinicians scrutinise the technology’s recommendations and check its accuracy before deciding the type of care and treatment a patient receives.
Secondly, the AI system can be easily applied to different types of eye scanner, not just the specific model on which it was trained. This could significantly increase the number of people who benefit from this technology and future-proof it, so it can still be used even as OCT scanners are upgraded or replaced over time.
The next step is for the research to go through clinical trials to explore how this technology might improve patient care in practice, and regulatory approval before it can be used in hospitals and other clinical settings. If clinical trials are successful in demonstrating that the technology can be used safely and effectively, Moorfields will be able to use an eventual, regulatory-approved product for free across all 30 of their UK hospitals and community clinics, for an initial period of five years.
Dr Pearse Keane graduated from UCD with an MB BCh BAO and a BMedSc in 2002 and he completed a MSc in Physiology in 2004. He was awarded an MD by the University in 2011 for his research under the supervision of Prof Colm O’Brien into the development of novel techniques in retinal imaging and their application to macular disease. He is a frequent visitor to Ireland and has acted as an external examiner in Ophthalmology for the School for several years. We congratulate Dr Keane and his colleagues on this success and wish them continued success as they seek to bring this approach into routine clinical practice.
We asked a number of our academic staff to comment on the significance of this paper. Please note none of these individuals are involved in this published research.
Prof Catherine Godson, Director of the UCD Diabetes Complications Research Centre observed,
There is a global epidemic of diabetes. It is currently estimated that there are 425 million people living with diabetes. This figure is predicted to rise to 700 million in the next decade. Diabetes is a leading cause of blindness. Earlier diagnosis will facilitate more effective and treatment and disease management. Given the prevalence of diabetes and its serious life limiting complications including kidney disease, heart disease and stroke applications of AI have huge potential benefit in earlier detection, understanding and disease management.
Prof Colm O’Brien, UCD Professor of Ophthalmology and consultant ophthalmologist at the Mater Misericordiae University Hospital commented,
These results are exciting for ‘screening’ large populations to detect sight threatening retinal diseases like diabetic retinopathy and macular degeneration – early detection and treatment where appropriate will mean that fewer patients will become visually impaired in the future – that’s the key take home message from this research paper.
The next challenge for the group will be to see if AI can help monitor patients with established disease in our clinics, and identify those at risk of progression. Also, this AI will also be very valuable at early detection of other ocular conditions such as glaucoma by examination of the optic nerve using the same OCT scans. Given the widespread use of high powered smartphone cameras, AI might well find further applications.
Mr David Keegan, UCD Clinical Professor in Ophthalmology and Retina at the Mater Misericordiae University Hospital, Mater Private Hospital and Temple Street Children’s University Hospital noted,
Pearse has become one of the world authorities on retinal imaging and his technical expertise and knowledge is well recognised. We hope to establish this type of infrastructure alongside the National Diabetic Retinal Screening Programme (RetinaScreen) in the future. The challenges are ensuring low false negatives (to ensure safety) but also to ensure low false positives to retain the integrity of the Programme and not swamp our treatment clinics with unnecessary referrals. We are making good steady progress in this regard since the Programme was established in 2013. As we have done with surveillance screening we will likely run a parallel pilot Programme using the best evidenced/positioned AI system.
De Fauw J, Ledsam JR, Romera-Paredes B, Nikolov S, Tomasev N, Blackwell S, et al. Clinically applicable deep learning for diagnosis and referral in retinal disease. Nature Medicine. 2018. [link]