Google–Moorfields Venture Eyes AI to Speed Diagnosis with OCT
Google’s DeepMind Health artificial intelligence project and Moorfields Eye Hospital in London have teamed up for a five-year venture to develop an algorithm using one million anonymized retina scans from the Moorfields’ database to speed up and improve diagnosis of age-related macular degeneration (AMD) and diabetic retinopathy.
Google and Moorfields announced the collaboration this month. Researchers said the algorithm will access a database of optical coherence tomography (OCT) scans and what Moorfields described as “some related anonymous information about eye condition and disease management” the hospital has collected.
Sir Peng Tee Khaw, MRCP, DO, director of the National Institute for Health Research Specialist Biomedical Research Centre in Ophthalmology at Moorfields, explained the rationale for the collaboration. “Our research with DeepMind has the potential to revolutionize the way professionals carry out eye tests and could lead to earlier detection and treatment of common eye diseases such as age-related macular degeneration,” he said. “With sight loss predicted to double by the year 2050, it is vital we explore the use of cutting-edge technology to prevent eye disease.”
The project came about after Moorfields ophthalmologist Pearse Keane, MD, approached DeepMind via its website to explore how DeepMind’s machine learning technology could be used to analyze scans to provide a better understanding of eye disease.
Mustafa Suleyman, co-founder of DeepMind, explained why DeepMind signed on to the collaboration with Moorfields. “We set up DeepMind because we wanted to use AI to help solve some of society’s biggest challenges, and diabetic retinopathy is the fastest growing cause of blindness worldwide,” he said.
Moorfields said its clinicians perform more than 3,000 OCT scans weekly. Managing such a large volume of data can cause delays in patient testing and in getting appointments to discuss diagnosis and treatment. The DeepMind–Moorfields collaboration aims to discover if by using diagnostic algorithms AI can remove this bottleneck to speed time to diagnosis and treatment.
Critics attacked the data-sharing agreement because individual patients had not given explicit consent. DeepMind answered: “It’s not possible to identify any individual patients from the scans.” Moorfields said patients may opt out of any data-sharing system by contacting the trust’s data protection officer.
DeepMind previously set up a similar project at the Royal Free, Barnet, and Chase Farm hospitals to use machine learning to analyze data for kidney disease, which drew criticism about data sharing.
This time, Moorfields has set up a Q&A on its website that, among other things, answers concerns about data protection.
SHARE THIS ARTICLE WITH YOUR COLLEAGUES:
RECOMMENDED FOR YOU
OIS@ASRS Reporter’s Notebook: Three Things Learned at OIS@ASRS About What Funders Want, Going Global, and AI