EEG Based Dynamic Functional Connectivity Analysis in Mental Workload Tasks With Different Types of Information

Guan, Kai; Zhang, Zhimin; Chai, Xiaoke; Tian, Zhikang; Liu, Tao; Niu, Haijun · 2022 · Crossref

DOI: 10.1109/tnsre.2022.3156546

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This study addresses the challenge of accurately evaluating mental workload (MWL) in cross-task scenarios, where traditional physiological metrics often fail due to varying information processing mechanisms across different task types. The authors propose a dynamic functional connectivity analysis method based on EEG microstates to identify "task-independent" neural markers. By focusing on the time-varying properties of brain networks, the research aims to overcome the limitations of static connectivity analyses and improve the robustness of MWL assessment in human-machine systems. The experimental design involved 16 healthy participants performing four distinct N-Back tasks manipulating verbal, object, spatial (verbal), and spatial (object) information at three workload levels. EEG data were recorded using a 60-channel system and pre-processed to remove artifacts. The authors employed the Topographic Atomize & Agglomerate Hierarchical Clustering (T-AAHC) algorithm to identify six global microstate classes (A–F). Dynamic functional connectivity networks were constructed for each microstate in the Theta and Alpha frequency bands using Phase-Locking Value (PLV). Graph theory metrics, including nodal degree, characteristic path length, clustering coefficient, and global efficiency, were calculated to quantify network topology. Statistical analyses included repeated-measures ANOVA and Pearson correlations, while a Support Vector Machine (SVM) classifier was used to validate the discriminative power of the selected features. The results demonstrated that 15 specific nodes and 68 connectivity pairs, primarily located in the Frontal-Parietal region, were sensitive to mental workload across all four task types. Specifically, the characteristic path length of the Microstate D brain network decreased, while global efficiency increased significantly in both Theta and Alpha bands as mental workload intensified. These changes indicate a trend toward higher functional integration and faster information transfer in the cognitive control network under high workload conditions. The SVM classifier achieved an average accuracy of 95.8% for within-task discrimination and 80.3% for cross-task discrimination, confirming the effectiveness of the proposed features. The study concludes that dynamic functional connectivity metrics derived from specific EEG microstates, particularly Microstate D, provide a robust basis for cross-task mental workload evaluation. The findings suggest that the brain’s cognitive control network exhibits consistent topological changes regardless of the type of information being processed. This approach offers a promising solution for developing accurate, objective, and continuous MWL monitoring systems in complex human-machine interaction environments, enhancing operator safety and task execution reliability.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-nvidia on 2026-08-10; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• neuro workload indices
• workload measurement

Information type

What kind of knowledge this paper contributes, grouped by family — independent of topic (what it is about) and method (how it was studied).

• Empirical Findings: physiological data