AI-based speech pattern analysis may allow Alzheimer's diagnosis by phone

Speech patterns in phone conversations can be used to identify people with early Alzheimer's disease, a new study suggests. Photo by Sabine van Erp/Pixabay
Speech patterns in phone conversations can be used to identify people with early Alzheimer's disease, a new study suggests. Photo by Sabine van Erp/Pixabay

July 14 (UPI) -- Patterns of speech in a phone conservation can be used to correctly identify adults in the early stages of Alzheimer's disease, a study published Wednesday by the journal PLOS found.

Using more than 1,600 voice recordings of phone conversations made from 24 people with confirmed Alzheimer's and 99 healthy controls, researchers correctly identified those with the common form of dementia with roughly 90% accuracy, the data showed.


The approach relies on the tendency of people with Alzheimer's "to speak more slowly and with longer pauses and to spend more time finding the correct word," the researchers said.

These "vocal features" result in "broken messages and lack of speech fluency," which can be analyzed using an artificial intelligence-based program.

RELATED Biogen revises label for new Alzheimer's drug, limiting use to those with mild dementia

The computer program was able to identify those with early Alzheimer's with essentially the same level of accuracy as a telephone-based test for cognitive function, according to the researchers.

"There is growing consensus that the presence of language deficits could be a part of clinical manifestation of Alzheimer's disease and mild cognitive impairment," researchers from several universities in Japan wrote in the study.

"[The] assessment of language production might be able to represent a unique opportunity for early detection of Alzheimer's disease," they said.


More than 6 million adults age 65 and older in the United States have Alzheimer's disease, the most common form of dementia, the Alzheimer's Association estimates.

The condition causes memory loss and declines in cognitive function, including the ability to perform activities of daily living, according to the association.

Alzheimer's can be challenging for physicians to detect at an early stage because those afflicted and their family members may not recognize the warning signs or be reluctant to share them, the association said.

RELATED Most cases of dementia in U.S. seniors go undiagnosed, study says

However, changes in speech patterns experienced by people in the early stages of the disease are increasingly being seen as a possible way identify it earlier and initiate treatment before it progresses, research indicates.

While no cure for Alzheimer's is available, medications can help slow its progression in some people, according to the Alzheimer's Association.

For this study, voice recordings were analyzed using an artificial intelligence-based computer software program designed to identify speech patterns typical of those in the early stages of the disease.

The accuracy of the program at identifying participants with early Alzheimer's was compared with that of the Telephone Interview for Cognitive Status, a widely used test in which subjects are asked questions and evaluated based on their responses and response times.


Both approaches successfully identified recordings of participants with Alzheimer's with 86% to 89% accuracy, the data showed.

"The findings of our study can create the opportunity for building new tools to identify Alzheimer's disease risk by using only vocal features obtained from daily conversations via telephone," the researchers wrote.

The approach could enable "early detection and diagnosis of dementia, including Alzheimer's disease and dementia in the sense that the tool can be used not only by healthcare professionals ... but also the general population at home," they wrote.

Latest Headlines


Follow Us