VALUE PROPOSITION
Early discrimination and prediction of mild cognitive impairment (MCI) are crucial for the study of neurodegenerative mechanisms and interventions to promote cognitive resiliency. Existing work mainly focuses on the detection of MCI before Alzheimer’s onset; however, quantitative methods which enable prediction of progression trends in older adults, especially pre-MCI diagnosis, are far less frequent. Here, focusing on resting-state EEG data collected among high-risk, under-served African American seniors, we first conduct multi-scale analysis of brain functional connectivity and obtain a series of discrimination approaches for healthy control (HC) and MCI, then combine diversified approaches to quantify the state of cognitive health. The novel EEG-based discrimination model demonstrates high sensitivity and stability for MCI detection. In addition, each decision on HC or MCI also comes with an EEG-based cognitive status score, which shows promising capability in predicting the personal progression trend of cognitive health in older adults, especially African American seniors. The present system allows early detection and prediction of people at risk of mild cognitive impairment before clinical symptoms may occur.
DESCRIPTION OF TECHNOLOGY
This technology uses resting-state EEG data collected among high-risk seniors. The EEG signals are processed to generate artifact reduced signals, then generating current source density signals from the artifact reduced signals. A multiscale analysis of dynamic functional connectivity of a brain is determined based on the current source density signals using multiple time windows. Hard classifiers are generated for each of the time windows. The system provides discrimination of normal cognition and mild cognitive impairment using selected classifiers. An EEG-based health score of overall brain activity is created based on majority voting of the selected classifiers. A display corresponding to the health score of brain activity is generated.
BENEFITS
APPLICATIONS
IP Status
US Patent Pending
LICENSING RIGHTS AVAILABLE
Full licensing rights available
Inventors: Tongtong Li, Jinxian Deng, Boxin Sun, Bruno Giordani, Mingyan Liu, Voyko Kavcic
Tech ID: TEC2023-0116
For more information about this technology,
Contact Raymond DeVito, Ph.D. CLP at Devitora@msu.edu or +1-517-884-1658