Self-Regulation Phenomenon Emerged During Prolonged Fatigue Driving: An EEG Connectivity Study

Li, Gang; Wang, Jie; Xu, Wanxiu; Wu, Kuijun; Liu, Yufeng; Bezerianos, Anastasios; Sun, Yu · 2023 · Crossref

DOI: 10.1109/tnsre.2023.3339768

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 investigates the neural mechanisms underlying driving fatigue, specifically addressing the "fatigue self-regulation" (FSR) phenomenon. While driving fatigue is a major cause of traffic accidents, previous research often relied on binary classification (vigilant vs. fatigued) or assumed monotonic deterioration in performance. However, some studies have observed non-monotonic trends in behavioral and neurophysiological indicators. The authors hypothesize that these fluctuations represent an adaptive self-regulation process where the brain temporarily compensates for fatigue. The study aims to quantify this phenomenon using electroencephalography (EEG) connectivity and behavioral data to distinguish between drivers who exhibit self-regulation and those who do not. The experimental design involved 26 healthy university students performing a 90-minute simulated driving task. EEG data were recorded using a 24-channel wireless dry electrode cap, and reaction times (RT) were measured as participants responded to brake lights of a guide car. To identify the FSR phenomenon, the researchers employed a data-driven clustering approach. RTs were normalized and clustered into five levels using K-means clustering. Subjects were categorized into an FSR group if their RT trajectory showed a non-monotonic trend with drops of more than two levels, indicating a temporary improvement in performance. The FSR group was further divided into four states: vigilant state (VS), fatigue state prior to regulation (FS), fatigue self-regulation state (FSRS), and fatigue deepening state (FDS). The non-FSR group, exhibiting monotonic RT increases, was divided only into VS and FDS. Functional connectivity (FC) was assessed using the phase lag index (PLI), and brain network characteristics were analyzed using graph theory metrics, including clustering coefficient, characteristic path length, and global/local efficiency. The results revealed significant differences between the FSR and non-FSR groups in both behavioral performance and brain network topology. The FSR group exhibited the predicted non-monotonic behavioral trend (increasing-decreasing-increasing RT), whereas the non-FSR group showed continuous deterioration. Neurophysiologically, the FSR group demonstrated distinct functional connectivity patterns and network reorganization during the FSRS phase compared to the non-FSR group. Specifically, the brain networks of the FSR group showed adaptive changes in integration and segregation metrics, suggesting a compensatory mechanism. Classification models using support vector machines, k-nearest neighbors, and BP Adaboost achieved high accuracy in distinguishing between the groups and states, validating the distinct neural signatures of the FSR phenomenon. The significance of this work lies in its challenge to the traditional view of fatigue as a linear accumulation process. By identifying and characterizing the FSR phenomenon, the study provides new insights into the complex, adaptive neural mechanisms of driving fatigue. These findings suggest that individual differences in fatigue resilience are linked to the brain's ability to self-regulate. This understanding can inform the development of more accurate, personalized fatigue detection systems and mitigation strategies that account for dynamic changes in driver state, rather than relying on static thresholds.

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.

• drowsiness detection algorithms
• time on task
• drowsiness
• neuro workload indices
• vigilance
• truck driver fatigue

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