AI can identify Alzheimer’s disease a decade before symptoms appear

An algorithm has reached more than 80 per cent accuracy in identifying structural changes to the brain caused by Alzheimer’s
An algorithm has reached more than 80 per cent accuracy in identifying structural changes to the brain caused by Alzheimer’s
RAFE SWAN/GETTY IMAGES

Artificial intelligence can identify the signs of Alzheimer’s in scans a decade before symptoms show, scientists say.

Researchers have developed an algorithm that analyses MRI scans to identify structural changes to the brain caused by the disease, with accuracy of more than 80 per cent. Experts believe that AI could be used by the NHS to predict Alzheimer’s and other diseases within ten years.

It is expected to be available sooner privately for those worried about mild memory loss, often the first sign of dementia, or family history. Having a close relative with the disease is thought to raise slightly the risk of developing it.

AI algorithms can make predictions without being explicitly programmed after they are “trained” or fed information from a data set.