Related Experiment Video
Updated: Jan 12, 2026
06:40
Microstate and Omega Complexity Analyses of the Resting-state Electroencephalography
Published on: June 15, 2018
10.6K
Reliability and state-dependency of EEG connectivity, complexity and network characteristics.
L S Dominicus1, D Y Lodema2, B Oranje3
1Department of Psychiatry, University Medical Center Utrecht, Utrecht, The Netherlands. l.s.dominicus-2@umcutrecht.nl.
Scientific Reports
|November 4, 2025
Summary
Resting-state electroencephalography (EEG) metrics like permutation entropy (PE) and phase-lag index (PLI) in theta and alpha bands show high reliability for biomarker development. Other metrics, such as minimum spanning trees (MST) and amplitude-envelope correlation (AECc), require caution due to instability.
Area of Science:
- Neuroscience
- Biomarker Development
- Signal Processing
Background:
- Resting-state electroencephalography (EEG) metrics are sensitive to task instructions and cognitive state, limiting their utility as stable biomarkers.
- Understanding the reliability of different EEG metrics across time and cognitive states is crucial for developing robust neurophysiological markers.
Purpose of the Study:
- To assess the temporal and state-dependent stability of three classes of resting-state EEG metrics: functional connectivity (FC), signal complexity, and network topology.
- To identify reliable EEG metrics suitable for biomarker development in healthy adults.
Main Methods:
- Sixty-four-channel EEG data were collected from healthy adults during two separate resting-state sessions and during semi-resting-state epochs within a task.
- Reliability was quantified using intraclass correlation coefficients (ICC) for functional connectivity (amplitude-envelope correlation, AECc; phase-lag index, PLI), signal complexity (permutation entropy, PE), and minimum spanning tree (MST) metrics.
- Analyses were performed at sensor and source levels, focusing on delta, theta, alpha, and beta frequency bands.
Main Results:
- Permutation entropy (PE) demonstrated consistently good-to-excellent reliability (ICC > 0.75-0.90) across sessions and states.
- Functional connectivity (FC) and minimum spanning tree (MST) metrics exhibited variable reliability, ranging from poor to good.
- Theta and alpha frequency bands generally showed higher reliability than delta and beta bands for most metrics. Alpha and theta PE, and alpha PLI, were identified as the most robust measures.
Conclusions:
- Theta and alpha permutation entropy (PE) and alpha phase-lag index (PLI) are robust and suitable for resting-state EEG biomarker development.
- Minimum spanning tree (MST) and amplitude-envelope correlation (AECc) metrics should be used with caution due to their limited stability across time and cognitive states, particularly outside the theta and alpha bands.
