AI could spot Alzheimer’s risk years in advance

LONDON: Scientists claimed an exciting breakthrough in treating Alzheimer’s after using artificial intelligence to predict who will get the disease for the first time.

Using the latest machine learning techniques, researchers successfully forecast which patients would go on develop Alzheimer’s up to eight years in advance.

But they cautioned that the study was very small – involving just 40 people – and that much more work is needed to test and refine the method.

The researchers created an algorithm by feeding more than 2,000 brain scans belonging to 1,000
patients through a deep learning tool so that it could recognise patterns associated with cognitive
decline.

They then applied the algorithm to brain scans of another 40 patients and identified those who later went on to develop Alzheimer’s.

“We’re very excited by the algorithm’s performance. It was able to predict every single case that advanced to Alzheimer’s Disease,” said Jae Ho Sohn, of the University of California in Berkeley.

He said that in a best case scenario, the algorithm could start appearing in the health service in three years.

“Early prediction is a hot topic because it opens up the possibility to intervene with therapy before irreversible damage occurs in the brain. If we only diagnose Alzheimer’s when all the symptoms have appeared, the brain volume loss is so significant that it’s too late to intervene,” he added.

Dr Sohn believes this is the rest time AI has been used to accurately determine a patient’s Alzheimer’s
risk.

Experts not involved in the research welcomed the findings – while cautioning that further studies are
needed.

Prof John Hardy, of University College London, said: “This is a tiny data set, only looking at 40 people. It’s also a very selected data set and not representative of the whole population. So we can’t know yet whether this is relevant to most people.”