High-density EEG network analysis in MCI: an exploratory study of electrode density and cognitive performance

Area of Science:

• Neuroscience
• Cognitive Science
• Biomedical Engineering

Background:

• Mild cognitive impairment (MCI) is associated with cognitive dysfunction.
• Electroencephalography (EEG)-derived functional connectivity and graph-theoretical measures are explored as potential biomarkers for MCI.
• The impact of electrode density on the clinical sensitivity of these EEG measures remains unclear.

Purpose of the Study:

• To investigate if source-level network metrics derived from EEG are associated with cognitive performance in individuals with MCI.
• To determine if electrode density influences the ability of these metrics to capture and classify cognitive differences in MCI.

Main Methods:

• Resting-state high-density EEG was recorded from 17 individuals with MCI.
• Source-level lagged linear connectivity and graph-theoretical metrics were computed from 173, 64, and 18-channel EEG montages.
• Associations with cognitive scores were examined, between-group differences were assessed, and classification performance was evaluated.

Main Results:

• Correlation patterns became clearer with increasing electrode density, particularly in alpha and beta bands.
• The 173-channel montage demonstrated the most significant between-group differences and best classification performance (balanced accuracy up to 0.82, AUC up to 0.94).
• These results suggest higher electrode density yields more sensitive network metrics.

Conclusions:

• Electrode density significantly impacts the sensitivity of source-level EEG network analysis for detecting cognition-related alterations in MCI.
• High-density EEG recordings provide richer network information, improving the ability to differentiate cognitive status within the MCI spectrum.