Recognizing mild cognitive impairment based on network connectivity analysis of resting EEG with zero reference.

Area of Science:

• Neuroscience
• Medical Imaging
• Biomedical Engineering

Background:

• Mild cognitive impairment (MCI) is a critical precursor to Alzheimer's disease (AD), necessitating early diagnostic tools.
• Disorders in multiple brain areas are associated with MCI, highlighting the need for network-based analytical approaches.
• Electroencephalography (EEG) offers a non-invasive method for assessing brain activity and network dynamics.

Purpose of the Study:

• To evaluate the efficacy of resting-state EEG-based brain network analysis for reliable MCI detection.
• To investigate the impact of different EEG reference techniques on MCI classification accuracy.
• To compare network-based EEG analysis with traditional coherence-based methods for MCI differentiation.

Main Methods:

• Analysis of resting-state EEG data across various frequency bands to construct brain networks.
• Systematic evaluation of different EEG reference choices, including the Reference Electrode Standardization Technique (REST).
• Comparison of classification performance between network-based analysis and coherence-based methods for MCI detection.

Main Results:

• Network-based MCI differentiation demonstrated superior performance compared to traditional EEG coherence methods.
• The choice of EEG reference significantly influenced classification accuracy.
• The Reference Electrode Standardization Technique (REST) enabled the construction of more accurate scalp EEG networks, leading to enhanced MCI differentiation.

Conclusions:

• Resting-state EEG network analysis is a promising tool for the early recognition of MCI.
• Standardizing EEG reference techniques, such as REST, is crucial for optimizing diagnostic accuracy.
• This approach holds potential for future clinical applications in Alzheimer's disease diagnostics and intervention.