An artificial intelligence (AI) tool has identified differences in the shape and size of blood vessels at the back of the eye, which correspond to cognitive status, according to research led by scientists at City St George’s, University of London.
Published
The study, funded by Alzheimer’s Research UK, unveils findings that could bring opticians one step closer to spotting warning signs of memory-stealing conditions like dementia and Alzheimer’s years earlier than currently possible.
The blood vessels in the eye act as an extension of those in the brain. This gives researchers the unique opportunity to analyse photos of the retina as a non-invasive way to gain insight about the characteristics of the blood vessels elsewhere, as a marker of brain health and signs of neurodegeneration.
In the largest scale peer-reviewed study of its kind, researchers looked at 110,282 eye scans from over 63,000 people aged 40 to 69 who were enrolled into the UK Biobank between 2006 and 2010.
Detecting changes in seconds
They applied the AI tool ‘Quartz’, developed by scientists at City St George’s and Kingston University, to the images to quickly and automatically detect the different types of small blood vessels (arterioles and venules) of the retina in just five seconds. The technology measured their width, area and the extent to which they twisted and turned.
The team compared these blood vessel measurements to the cognitive score of individuals when they entered the study. Cognitive status was assessed using a score determined by a combination of four tests that assessed two types of memory, reaction time and intelligence.
They found that a lower cognition score – indicative of a decline in thinking ability – was associated with reductions in arteriole and venular width and area, reduced arteriolar twisting and increased venular twisting.
These associations were adjusted for age, sex, ethnicity and UK Biobank centre, and were independent from the patterns found with age alone.
The results suggest that there are different vascular processes linked to cognition. The researchers say that these could relate to a reduced blood supply, which might be an early cause or consequence of neurodegenerative disease.
Seamlessly embed into daily optician routines
Professor Chris Owen, who is leading the study, now plans to use this Quartz AI tool in other studies and populations to see how it could be used more widely in the healthcare sector.
The team will then need to build prediction models, which examine how these measures can be used, to predict future cognitive decline and risk of neurodegenerative disease.
Professor Chris Owen, lead author and Professor of Epidemiology at City St George’s, University of London said:
“There’s huge scope for our AI technology to be seamlessly embedded into the day-to-day workflows of opticians on the high-street and in eye clinics."
"It would open-up a new avenue to get those who need preventative treatment much sooner to avert diseases like Alzheimer’s and dementia from taking hold altogether or delay it in its tracks."
A gamechanger to spot dementia early
David Thomas, Head of Policy and Public Affairs at Alzheimer’s Research UK said:
“We’ve seen incredible progress in new diagnostic tools, like blood tests, in recent years, and eye scans seem to be following the same promising path.
Right now, people are waiting too long for a dementia diagnosis, often struggling to even get through the NHS front door. Detecting early signs of dementia, for example through a routine eye test at your local optician, would be a gamechanger. Integrating these scans with existing tests, like brain scans, would ensure people get the full picture of their condition, and the support they deserve, more quickly."
The study was published in the journal Alzheimer's & Dementia: Diagnosis, Assessment & Disease Monitoring.