Developing AI technology to detect the first signs of Alzheimer’s disease

With support from a new $ 3.9 million grant, Michigan State University researchers are developing technology that analyzes speech and vocabulary patterns for early signs of Alzheimer’s disease , the most common cause of dementia.

Jiayu Zhou, associate professor at the MSU College of Engineering, is leading the effort powered by artificial intelligence, or AI, and funded by the National Institutes of Health. In collaboration with Oregon Health & Science University and Weill Cornell Medicine, the goal is to code an easy-to-use smartphone app to help assess whether a follow-up medical diagnosis is needed.

“Alzheimer’s disease is difficult to manage and it’s very easy to confuse its early stage, mild cognitive impairment, with normal cognitive decline as we get older,” said Zhou, who heads a research group in the Department. computer science and engineering. “It’s only when things get worse that we realize what’s going on and, by then, it’s too late.”

Although there is currently no cure for Alzheimer’s disease, catching it earlier could help doctors and researchers develop a treatment to slow it down or stop it before it causes irreparable damage. .

And Zhou believes AI can detect more subtle changes in speech and behavior earlier and more reliably than human observers. Additionally, bundling the power of AI into one app would make it much more affordable and accessible than medical diagnostics, such as MRIs and in vivo testing. These tests can be lengthy, invasive and extremely expensive, Zhou said.

While this approach to AI may sound like science fiction, Zhou and his team have already shown in preliminary tests that it is as accurate as MRIs at recognizing warning signs. The tests used data collected by collaborators at Oregon Health & Science University who are leading a clinical trial investigating how conversations could be used as a therapeutic intervention for dementia or early Alzheimer’s disease.

These trials gave the Spartan team hours and hours of interviews that they could use to test their AI. The interviews were transcribed and the algorithm was able to scan the text, looking at things like the variety of words people were using to assess their cognitive status.

The team’s initial success with this data prompted them to pursue this new grant and support their project in two major ways. The first is to reduce the time it takes for the AI ​​to do an assessment.

“If we want to develop an app that everyone can use, we don’t want people talking to them for hours,” Zhou said. “We need to develop an effective strategy so that we can navigate the conversation and get the data we need as quickly as possible, in 5-10 minutes.”

The team already has a prototype app that queries a user and records their audio responses. One of the next steps is to refine the questions the app asks, and how it asks them, to get what it needs from users faster.

The second goal is to bring data beyond linguistic models that will help AI to make an assessment. For example, the application will also examine the acoustic signals of the conversation and can also make more use of the video to analyze facial expressions as well as the words spoken by the user. The team is also working on integrating behavioral sensors that would track things like a person’s amount of sleep to complement the app’s interview language analysis.

The app would digest all of that data and then give users risk scores indicating how likely they are to show signs of dementia. Zhou pointed out that ultimately, however, it would be a doctor – not a computer – who would make this diagnosis. Still, the AI-powered app would expand affordable self-assessment technology to help millions of people and could encourage patients to seek help sooner.

“You cannot replace this human interaction,” Zhou said. “The final assessment will be done by a patient’s doctor. But if you have any doubts and the app says you are more at risk, you don’t have to wait. You can see a clinician and move on to following steps.”

Hiroko Dodge, professor of neurology at Oregon Health & Science University, and Fei Wang, assistant professor of health care policy and research at Weill Cornell Medicine, will join Zhou for this grant.