Published on
Updated
Reading 2 mins.
in collaboration with
Dr Christophe de Jaeger (Longevity and geriatrics)
Medical validation:
July 13, 2022
This new tool, developed by researchers at Boston University, could enable early diagnosis of cognitive disorders from voice recordings. Update on this progress with Dr. Christophe de Jaeger, physiologist and specialist in aging.
The objective of this computer 2.0? Detecting cognitive disorders from audio recordings, without a physical appointment. An ambitious project, the results of which were published in the Journal of the Alzheimer’s Association.
Early diagnosis, a “major medical challenge”
In France, about 1 million people are affected by Alzheimer’s disease. Worldwide, there are nearly 50 million. This is why early diagnosis of the disease is absolutely essential.
“Early diagnosis of degenerative diseases is a major medical challenge because these pathologies are too often diagnosed late. In reality, individuals fear the diagnosis: they therefore do not do a test. This is a real mistake” says Dr. Christophe de Jaeger, physiologist and specialist in aging.
To detect the early stages of the disease, the scientists therefore developed an online model using audio recordings of neuropsychological tests. They listened to the sound recordings of more than 1,000 people and based themselves on one test in particular – the “Boston Naming Test” – which consists of putting a picture to a word.
As a result of the examinations, the tool proved not only able to accurately distinguish between healthy people and those with dementia, but also to differentiate between people with mild cognitive impairment and those with dementia, depending on the co- author of the article, Ioannis Paschalidis.
Another interesting point: the content of the recordings was more important than the way the volunteers expressed themselves.
A tool that does not replace “etiological research”
“It surprised us that the speech stream or other audio features weren’t so critical; you can pretty much automatically transcribe interviews and rely on AI-powered text analytics to assess cognitive impairment”assures the professor.
Clearly, this new method could help doctors assess patients who are at risk of developing some form of cognitive decline.
“Finding a way to make a diagnosis quickly and easily, with a good degree of reliability, seems to me to be the challenge of this beginning of the century”, points out Dr. Jaeger. According to him, “whatever the type of analysis or the duration of the test, I think that nothing will replace an etiological investigation, which consists in looking for the causes of an affection”.
For the physiologist and aging specialist, this computerized approach cannot replace a scanner.
“Biologically, a certain number of elements will be missed by clinicians with this tool. It must therefore be seen as an additional element to be offered for diagnosis, but not as a means of making a diagnosis or finding a treatment” tempers Dr. Christophe de Jaeger.