HomeProjectsInformation and Communication TechnologiesMachine Learning Technologies for Advancing Digital Biomarkers for Alzheimer’s Disease

Machine Learning Technologies for Advancing Digital Biomarkers for Alzheimer’s Disease

Print IconPrint the page
Early identification of people at risk of developing Alzheimer’s Disease (AD) and timely intervention to slow the onset and progression of AD are crucial. In this project, we will build an indoor monitoring system that deploys different types of sensors and collect the corresponding data to develop a multi-modal fusion algorithm for daily activities and behavior detection. Besides, we will develop a new federated learning framework, which not only ensures the real-time of related algorithms but also protects users’ private data. After that, we will predict and identify individuals who are at a higher chance of developing dementia and provide high-quality feedback to the users.
The illustration of the entire framework.
A demo of our fall detection system. (a): The man is standing. Fall has not happened. (b): Fall has happened. Our system can immediately give feedback to the caregiver so that the fallen people can be rescued in time.
Examples of detecting activities using motion features. The left column shows a dining scenario; the right column shows a TV viewing scenario.
Activities of daily living biomarkers and proposed detection algorithms.
Activities of daily living biomarkers and proposed detection algorithms.

Uniqueness and Competitive Advantages:

  • The collected data are stored in the local repositories and are neither being transmitted nor leaked, thus protecting the users’ privacy
  • Uses depth cameras to collect depth images of users’ daily activities, which don’t reveal face information
  • Provides users with the high-quality early identification of developing AD and timely intervention to slow the progression of AD

Do you like our project?

Contact us

MORE TO EXPLORE

All projects >