Research by Moorfields Eye Hospital, DeepMind Health and University College London (UCL) Institute of Ophthalmology in 2016 has highlighted that artificial intelligence (AI) can help to identify eye diseases just as accurately as optometrists (Optician, 17.08.18).
Their new system, which is due to start being used in clinical trials soon, was able to recommend the correct referral decision for more than 50 different eye diseases with 94% accuracy. Using two different mathematical systems, the technology can identify features of different eye diseases from OCT scans, before deciding how the patient should then be referred based on the urgency of the recognised conditions. As more than 285 million people worldwide live with a form of sight loss, including more than two million people in the UK, technology on this scale could be the future of optics.
With the introduction of this AI, doctors and other eye care professionals could spot any serious eye conditions earlier and prioritise the patients who need treatment more urgently.
Dr Pearse Keane, consultant ophthalmologist at Moorfields Eye Hospital NHS Foundation Trust and NIHR Clinician Scientist at the UCL Institute of Ophthalmology, said: ‘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.’
Alan Tinger, Foresight Report Steering Group chairman, said: ‘This is a very significant development for patient care and all credit to Pearse Keane and colleagues at Moorfields for their pioneering work.’
The Foresight report, published two years ago, stated that by 2030, there is likely to be a 50% rise in people aged over 65 and 100% more aged over 85. The average cost of providing hospital and community health services for a person at this age is around three times greater than for a person aged 65 to 74 years.
Increases to the population and indeed to the number of patients with potential eye diseases over the next few years will have an impact on the demand for optical services and products. As the population gets older, there may be multiple forms of eye disease and related health concerns, meaning that an AI system like this that can be attached to different types of eye scanners would be welcomed by optical professionals.
Moorfields has teamed up with Google’s DeepMind division, whose trained AI model can be used for any future non-commercial research projects of a similar nature.
Technology can help with the shortfall in eye care professional numbers
Mustafa Suleyman, co-founder and head of applied AI at DeepMind Health, said: ‘We set up DeepMind Health because we believe artificial intelligence can help solve some of society’s biggest health challenges, like avoidable sight loss. These incredibly exciting results take us one step closer to that goal and could, in time, transform the diagnosis, treatment and management of patients with sight threatening eye conditions, not just at Moorfields, but around the world.’
Professor Sir Peng Tee Khaw, director of the National Institute for Health Research (NIHR) Biomedical Research Centre at Moorfields Eye Hospital NHS Foundation Trust and UCL Institute of Ophthalmology, believed that the results of the research with DeepMind were exciting and demonstrate the potential sight-saving impact AI could have for patients. He said: ‘I am in no doubt that AI has a vital role to play in the future of healthcare, particularly when it comes to training and helping medical professionals so that patients benefit from vital treatment earlier than might previously have been possible.’
The Foresight report went on to discuss how ophthalmic technology such as this new AI system was likely to drive behaviour change among professionals, the public, manufacturers and businesses. With all the abilities that it has, the system can contribute positively to the UK’s eye care service delivery, it concluded.
Nicholas Rumney, senior optometrist at BBR Optometry in Herefordshire, said that this new technology is obvious and inevitable. He said: ‘It’s fascinating. However, it is not just making correct diagnoses, it is important that people who do not need referrals are not referred and that can only come from optometrists doing a better job at clinical decision making. Technology will certainly help that, but we need to be part of the system not outside the compound.’
Rumney said that it will benefit practices to have this kind of technology and that most practices already have something similar.
‘The Topcon Triton Swept-Source OCT produces a data set extracted in a template form called the Hood Report. This is from a single 12mm by 9mm scan encompassing optic disc and macular area,’ he added. ‘This has been shown to produce a positive capability of predicting glaucoma in the hands of naive users that is at least as good as expert glaucoma ophthalmologists.’
He also suggested that there was a need for more technological advances in optics and that they should be embraced in both clinics and business plans.
Sunil Shah, president of the BCLA, believed this technology was the way forward. He said: ‘We have a shortage of eye care professionals compared to the amount of pathology out there, so if AI can help reduce that burden and allow eye care professionals to concentrate on the patients that has to be good for everyone. AI has been used in screening for diabetic retinopathy from fundal photographs and this is the next step on.’
Shah believed that as the speed of advancement of technology was so fast, training within practices to interpret the findings needs to be able to keep up. ‘This sort of imaging is likely to undergo a secondary screening that perhaps could be directly linked to the practice images, giving an extremely rapid and effective decision-making process,’ he added.
As for a need for more technological advances in optics, Shah stressed that optical professionals must keep innovating. ‘Whether it is the transition from a photograph to the 3D OCT scan and beyond to being able to image easily on smart phones, we need to keep advancing. If not, we are not going to do the best for our patients.’