Predicting Alzheimer’s Disease Using AI
From brain imaging to blood tests and cognitive assessments, the tools available to detect Alzheimer’s disease generally catch this disease too late. But now, Boston University researchers are exploring the power of AI to predict Alzheimer’s by analyzing a patient’s speech.
Funded by the National Science Foundation, National Institute on Aging, and BU’s Rajen Kilachand Fund for Integrated Life Sciences and Engineering, the model they designed can predict, with an accuracy rate of 78.5%, whether someone with mild cognitive impairment is likely to remain stable or fall into the dementia associated with Alzheimer’s disease. The multidisciplinary team of engineers, neuropsychologists, and computer and data scientists published their findings in Alzheimer’s & Dementia: the Journal of the Alzheimer’s Association.
“We can reasonably make that prediction with relatively good confidence and accuracy,” says Ioannis Paschalidis, coauthor of the paper and director of the BU Rafik B. Hariri Institute for Computing and Computational Science & Engineering. “It shows the power of AI.”
The model was developed using data from the Framingham Heart Study, of which BU has been a longstanding partner. Using audio recordings of interviews with a group of study participants diagnosed with mild cognitive impairment, Paschalidis and his colleagues used speech recognition tools and machine learning to train their model to spot connections among speech, demographics (e.g., age, gender, etc.), diagnosis, and disease progression. Then, they tested the model’s predictive capabilities on the rest of the study’s participants. The results were impressive.
Their model may have the added benefit of providing healthcare equity for underserved populations. “Technology can overcome the bias of work that can only be done by those with resources, or care that has relied on specialized expertise that is not available to everyone,” says Rhoda Au, a professor of anatomy and neurobiology, and an investigator with the Framingham Heart Study, who coauthored the paper. With the hope to eventually eliminate expensive lab tests, imaging exams, or office visits, Au and Paschalidis see tremendous potential for diagnosing dementia from home, using a smartphone app.
“If you can predict what will happen, you have more of an opportunity and time window to intervene with drugs,” says Paschalidis. “We hope, as everyone does, that there will be more and more Alzheimer’s treatments made available.”