A new AI tool has been developed by a team of researchers at New Jersey’s Stevens Institute of Technology that can detect signs of Alzheimer’s in its early stages.

This research focused on how sufferers of Alzheimer’s express themselves in the form of speech. With differences in the language of sufferers having been discovered, the team then set out to produce an AI tool that could potentially be a new method of screening for the disease.

Alzheimer’s can damage areas of the brain that can control a person’s speech, which in turn leads to sufferers struggling to find the right words to say.

According to New Atlas, the team began their development of the AI tool by looking into ‘a standard picture description task’ – a popular exercise that is currently used in language screening for Alzheimer’s. During this, the subjects describe a drawing provided to them.

A transcription of the subject’s description is then used to train the team’s new AI algorithmic tool. The tool breaks down each of the sentences and labels them with numerical values, which enables the tool to ‘analyse the structural and thematic relationships’ between them. 

“We’re opening an exciting new field of research, and making it far easier to explain to patients why the A.I. came to the conclusion that it did, while diagnosing patients. This addresses the important question of trustability of A.I systems in the medical field,” said K.P. Subbalakshmi, who is the lead author of this study.

The team state that the tool has a 95% hit rate for distinguishing sentences spoken by both Alzheimer’s sufferers and healthy subjects. Looking forward, Subbalakshmi’s team are now aiming to optimise the tool to be used on different languages besides English.