Eisai harnesses wearables data for AI-led Alzheimer's prediction

Japanese pharmaceutical company Eisai, in collaboration with Oita University, has developed what could be the world's first AI model that utilises data from wearable devices to predict Alzheimer's disease. The research aimed to create a cost-effective tool for pre-screening individuals suspected of developing the disease. The study collected biological and lifestyle data from 122 individuals aged 65 and older, including physical activity, sleep, heart rate, and lifestyle factors. The researchers developed a predictive model using three machine learning technologies to assess the likelihood of participants testing positive for amyloid-beta protein accumulation, a crucial biomarker of Alzheimer's.

The AI model demonstrated "sufficient" capability to predict amyloid-beta protein accumulation and identified 22 contributing factors, such as physical activity, sleep, heart rate, age, education, and social engagement. Currently, Alzheimer's testing involves expensive and invasive methods like amyloid PET and cerebrospinal fluid testing. The Eisai and Oita University team's approach offers a promising alternative, focusing on lifestyle and biological data to make screening more accessible, particularly in an ageing population like Japan.

This research aligns with a broader trend of utilising wearable devices and AI for early detection of neurodegenerative diseases. The team envisions broader applications for their AI model in pre-screening individuals for Alzheimer's, especially those with limited access to current testing methods. Eisai continues its involvement in Alzheimer's research through its digital health business, Theoria, which focuses on developing a risk prediction algorithm for the early detection of mild cognitive impairment.

Click here to read the original news story.