It takes a lot of time – and money – to diagnose Alzheimer’s disease. After conducting lengthy in-person neuropsychological examinations, clinicians must transcribe, review, and analyze each response in detail. But researchers at Boston University have developed a new tool that could automate the process and possibly allow it to move online. Their machine-learning-based computer model can detect cognitive impairment from audio recordings of neuropsychological tests — no in-person appointment necessary. Their findings were published in Alzheimer’s and dementia: the journal of the Alzheimer’s Association.
“This approach brings us closer to early intervention,” says Ioannis Paschalidis, co-author of the paper and Emeritus Professor of Engineering at BU College of Engineering. He says that faster and earlier detection of Alzheimer’s disease could lead to larger clinical trials that focus on individuals in the early stages of the disease and potentially enable clinical interventions that slow cognitive decline: ” This can form the basis of an online tool that could reach everyone and could increase the number of people who get tested early.”
The research team trained their model using audio recordings of neuropsychological interviews of more than 1,000 people as part of the Framingham Heart Study, a long-running UB-led project looking at cardiovascular diseases and other physiological conditions. Using automated online speech recognition tools, think “Hey, Google!” – and a machine learning technique called natural language processing that helps computers understand text, their program transcribed interviews and then encoded them into numbers. A final model was trained to assess the likelihood and severity of an individual’s cognitive impairment using demographic data, text encodings, and actual diagnoses from neurologists and neuropsychologists.
Paschalidis says the model was not only able to accurately distinguish between healthy people and those with dementia, but also detect differences between people with mild cognitive impairment and dementia. And, it turned out that the quality of the recordings and the way people spoke – whether their speech was rapid or consistently hesitant – was less important than the content of what they said.
“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” , says Paschalidis, who is also the new manager of BU’s Rafik B. Hariri Institute for Computing and Computational Science and Engineering. Although the team has yet to validate their findings against other data sources, the results suggest that their tool could help clinicians diagnose cognitive impairment using audio recordings, including those of appointments. you virtual or telehealth.
Screening before symptoms appear
The model also provides insight into which parts of the neuropsychological exam might be more important than others in determining whether an individual has cognitive impairment. The researchers’ model divides exam transcripts into different sections based on the clinical tests performed. They found, for example, that the Boston Naming Test — in which clinicians ask individuals to label a picture using a word — is the most informative for an accurate diagnosis of dementia. “This could allow clinicians to allocate resources in a way that allows them to do more screening, even before symptoms appear,” says Paschalidis.
Diagnosing dementia early is not only important so that patients and their caregivers can create an effective treatment and support plan, but it is also crucial for researchers working on therapies to slow and prevent disease progression. Alzheimer’s. “Our models can help clinicians assess patients for their risk of cognitive decline,” says Paschalidis, “and then best tailor resources by performing additional testing on those with a higher likelihood of dementia.”
Interested in joining the research effort?
The research team is looking for volunteers to complete an online survey and submit an anonymous cognitive test. The results will be used to provide personalized cognitive assessments and will also help the team refine their AI model.
Source of the story:
Material provided by Boston University. Original written by Gina Mantica. Note: Content may be edited for style and length.