Feeling forgetful? AI may know if you’ll get Alzheimer’s

Scientists have used artificial intelligence to successfully predict if and when at-risk patients’ will develop Alzheimer’s disease. So would you want to know?

Cambridge University scientists have used artificial intelligence to successfully predict if and when at-risk patients’ will develop Alzheimer’s in four out of five cases.

Their algorithm analysed the cognitive tests and MRI scans of patients with early warning signs of dementia and was three times as accurate as the current clinical standards.

Research, published in eClinical Medicine, shows it was correct in 80 per cent of cases, with the team hoping it can reduce the need for invasive tests and the number of people diagnosed when it is too late for treatment.

Cambridge Psychology Professor Zoe Kourtzi said the tool didn’t just predict someone’s risk of developing the disease, but also whether “this progress will be fast or slow”.

“We’ve created a tool, which despite using only data from cognitive tests and MRI scans, is much more sensitive than current approaches at predicting whether someone will progress from mild symptoms to Alzheimer’s,” she said.

“This has the potential to significantly improve patient wellbeing, showing us which people need closest care, while removing the anxiety for those patients we predict will remain stable.”

She said artificial intelligence was only as good as the data it was trained on, and theirs had been “trained and tested” on “routinely-collected data not just from research cohorts, but from patients in actual memory clinics.”

he algorithm analysed the cognitive tests and MRI scans of patients with early warning signs of dementia and was three times as accurate as the current clinical standards.

The algorithm was built on the test results from 400 patients in a US study, before being tested on a larger group, including 900 Singapore and UK memory clinic patients whose symptoms had been tracked for up to six years.

It found about half of people who visited memory clinics concerned about mild cognitive impairment remained stable.

The other half developed Alzheimer’s — but more than half of these patients progressed slowly.

30 Professor Kourtzi said we needed tools that can better identify people in need of intervention if we want to tackle dementia.

“Our vision is to scale up our AI tool to help clinicians assign the right person at the right time to the right diagnostic and treatment pathway,” she said.

“Our tool can help match the right patients to clinical trials, accelerating new drug discovery for disease modifying treatments.”

Cambridge psychiatry assistant professor Dr Ben Underwood said it was exciting to know they might be able to reduce the uncertainty of a dementia diagnosis “with information we already have”.

“(It) is likely to become even more important as new treatments emerge,” he said.