[Research on classification of brain functional network features during mental fatigue]

Area of Science:

• Neuroscience
• Complex Systems Science
• Computational Psychiatry

Context:

• Mental fatigue, often induced by sustained cognitive tasks, poses challenges for accurate objective assessment.
• Traditional evaluation methods may lack precision, necessitating novel approaches.
• Brain functional networks offer a potential avenue for understanding cognitive states.

Purpose:

• To investigate objective indicators for evaluating mental fatigue using electroencephalogram (EEG) data.
• To enhance the accuracy of mental fatigue assessment through network analysis.
• To differentiate between normal and mental fatigue states using brain functional network characteristics.

Summary:

• Brain functional networks were constructed from EEG data during normal and mental fatigue states induced by cognitive tasks.
• Complex network theory was applied to analyze nodal characteristic parameters (degree, betweenness centrality, clustering coefficient, average path length).
• Support vector machine (SVM) models, optimized via grid search and cross-validation, classified subjects based on these network parameters, demonstrating successful differentiation.

Impact:

• Node characteristic parameters of brain functional networks can effectively distinguish between normal and mental fatigue states.
• This study provides a foundation for developing objective, EEG-based methods for mental fatigue evaluation.
• The findings contribute to a deeper understanding of the neural underpinnings of mental fatigue and its detection.