EEG Connectivity Signatures in Active vs. Passive Mental Fatigue Settings

Area of Science:

• Neuroscience
• Cognitive Science
• Occupational Health

Background:

• Mental fatigue in occupational settings arises from sleep deprivation and workload.
• Existing lab studies often fail to replicate real-world fatigue conditions.
• Biomarker validity for cognitive fatigue in active settings is underexplored.

Purpose of the Study:

• Investigate resting-state EEG functional connectivity in active and passive mental fatigue.
• Compare within-frequency (PLI) and cross-frequency (PAC) connectivity metrics.
• Assess transferability of fatigue biomarkers across different induction methods.

Main Methods:

• Utilized multilayer graph representation of sensor-space EEG data.
• Quantified intra-frequency (PLI) and inter-frequency (PAC) connectivity.
• Employed machine learning classifiers for feature validation and network topology analysis.

Main Results:

• Inter-frequency PAC showed higher informativeness for fatigue discrimination than intra-frequency PLI.
• Active fatigue settings required smaller feature sets for information saturation compared to passive settings.
• Neural resource reorganization varied significantly based on fatigue induction method.

Conclusions:

• Cross-frequency PAC is a promising biomarker for mental fatigue, applicable across diverse settings.
• Fatigue biomarker discovery requires consideration of the induction method's nature.
• Findings support the transferability of EEG-based fatigue biomarkers from lab to real-world conditions.