Using articificial intelligence to improve dementia diagnosis

25 April 22

written by:

Artificial intelligence that will analyse up to 6.3million eye scans could help improve the diagnosis of dementia and detect a range of other diseases.

A unique and specialised software which processes the images could be instrumental in identifying patterns of change in the retina associated with dementia, potentially a number of years before the brain starts showing any signs of change.

The information is taken and processed from AlzEye, a study that contains retinal images and ophthalmic data from thousands of patients.

The study led by UCL’s Dr Pearse Keane -one of the Top 100 most influential people in ophthalmology- will use this tool to ultimately improve the diagnosis of dementia, estimated to affect one million people in the UK by 2025.

The project could have even wider-reaching impact as researchers are also investigating the relationship between retina changes in a range of other systemic diseases.

The power of technology

Given the rich data on AlzEye, there have been numerous additional discoveries through the Fight for Sight funded project.

Researchers have found retinal blood vessel change across various other diseases (e.g. schizophrenia and multiple sclerosis) and risk factors (e.g. obesity).

The team has also explored the relationship of cardiovascular disease and the eye and found retinal anatomy to be a powerful indicator of incident heart attack and stroke.

Speaking about the project, Dr Pearse Keane told us:

“A recurring theme when presenting this research to the public has been a newfound motivation to attend their high street opticians more regularly. They tell us ‘I didn’t realise you could see diseases of the body in the back of the eye’. So this research could have far-reaching impact in not only detecting changes associated with dementia, but also encouraging the public to undergo eye checks more frequently, leading to earlier identification of many ophthalmic diseases.”

Long-term ambition

Over the next 12 months, data analysis will be completed on the key objectives of this project.

AlzEye will also require further funding to fully establish it as a bioresource with wider access to approved researchers.

The team’s long-term ambition is to provide a rapid imaging-based systemic health evaluation (particularly for cardiovascular diseases and dementia) from retinal images acquired regularly as part of someone’s routine care in hospital and ultimately, in a high street optician.