Topological EEG-Based Functional Connectivity Analysis for Mental Workload State Recognition

Yan, Yan; Ma, Liang; Liu, Yu-Shi; Ivanov, Kamen; Wang, Jia-Hong; Xiong, Jing; Li, Ang; He, Yini; Wang, Lei · 2023 · Crossref

DOI: 10.1109/tim.2023.3265114

archive: archived pipeline: cataloged

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

Summary

**Research Problem and Motivation** Mental workload (MWL) assessment is critical for preventing health issues and accidents in high-stakes environments like driving and aviation. While electroencephalography (EEG) is a standard tool for physiological MWL monitoring, existing methods often rely on traditional functional connectivity (FC) metrics or deep learning frameworks. This paper addresses the gap in utilizing topological data analysis (TDA) for EEG-based MWL recognition. Specifically, it investigates whether topological features derived from brain FC networks can effectively distinguish between different cognitive states, offering a novel alternative to conventional graph-theoretic descriptors. **Methodology** The authors propose a framework that extracts topological features from EEG-based FC networks using persistent homology, a technique from TDA. The process involves constructing FC networks using Pearson’s correlation coefficients (PCC) from multi-lead EEG signals. These networks are treated as graphs, and their topological properties are analyzed via graph filtration. The method tracks the birth and death of topological objects (connected components and voids) as the filtration parameter increases, generating Betti numbers and barcode summaries that characterize the network’s structure. The approach was validated on three public benchmark datasets: STEW (Simultaneous Capacity tasks), PhysioNet-MAT (mental arithmetic), and WAUC (Multi-Attribute Task Battery II with physical activity). Five specific classification tasks were defined, including binary distinctions between resting and working states, subjective workload ratings, and low versus high MWL levels during physical exertion. **Results** Experimental results across the three datasets demonstrate that the proposed topological FC analysis scheme exhibits excellent distinguishing ability for brain state recognition. The method performed comparably to or better than state-of-the-art results with similar settings. The topological features proved to be effective and robust indicators of brain states, successfully differentiating between low, moderate, and high workload levels, as well as between pre-task and during-task states in arithmetic and multitasking scenarios. **Significance** This work represents the first investigation of EEG-based MWL evaluation using persistent homology analysis of multivariate time series. By leveraging topological invariants, the study provides a new perspective on understanding brain network organization during cognitive tasks. The findings suggest that topological features are a viable and robust alternative for designing novel brain–computer interface systems, potentially enhancing the accuracy and interpretability of fatigue and workload monitoring in human-machine interaction contexts.

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 qwen3.8-27b-gittensor on 2026-08-23; verification: pending re-verification.

Topics

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

• neuro workload indices

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