EEG Features for Driver’s Mental Fatigue Detection: A Preliminary Work

Kamaruzzaman, Muhammad ‘Afiq Ammar; Othman, Marini; Hassan, Raini; Abdul Rahman, Abdul Wahab; Mahri, Nurhafizah · 2023 · Crossref

DOI: 10.31436/ijpcc.v9i1.355

archive: archived pipeline: cataloged verified

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

Summary

This paper addresses the critical public health issue of traffic accidents caused by driver mental fatigue, a condition resulting from cognitive overload and sleep deprivation that impairs attention, decision-making, and reaction times. While physical fatigue is well-documented, the authors argue that mental fatigue is a distinct and significant contributor to road safety risks, often overlooked in favor of physical exhaustion. The study aims to characterize mental fatigue by reviewing existing literature on electroencephalogram (EEG) features, specifically focusing on how neural activity changes during driving tasks. The motivation is to identify reliable EEG biomarkers that can be used to develop systems for detecting driver fatigue in real-time, thereby enhancing traffic safety. The authors employed a narrative review approach to analyze previous studies on EEG-based mental fatigue detection. They examined various methodological aspects, including EEG pre-processing techniques, feature extraction methods, and classification algorithms. The review covered qualitative methods like Event-Related Potentials (ERPs) and quantitative methods such as spectral analysis and functional connectivity analysis. Specific attention was given to studies utilizing different EEG channel configurations (ranging from 16 to 64 channels) and sampling rates. Additionally, the authors conducted a systematic search of three public data repositories—Kaggle, UCI Machine Learning, and Physiobank—to identify available datasets for future empirical research. The review found that spectral analysis is the most prevalent method for detecting mental fatigue, with consistent evidence pointing to an increase in alpha power, particularly in the parietal regions of the brain, as a key indicator of fatigue. Other studies highlighted increases in delta rhythm relative energy and decreases in sampling entropy. Functional connectivity analysis also suggested significant changes in the parietal and occipital areas. However, the search for public datasets yielded no results containing actual EEG data related to driving scenarios; the retrieved datasets consisted only of psychometric questionnaires regarding burnout and resource allocation. Consequently, the authors proposed a conceptual model for future research involving data acquisition using a 19-channel EEG device during various driving simulations, followed by signal pre-processing, feature extraction via Event-Related Potentials, and binary classification using logistic regression and Support Vector Machines. The significance of this work lies in its synthesis of current knowledge on EEG markers for mental fatigue, highlighting the parietal alpha band as a promising feature for detection. The authors conclude that while spectral analysis is effective and interpretable, generalization across different study conditions remains challenging due to variations in experimental design and sample sizes. The lack of public EEG datasets underscores the need for new data collection efforts. Future work should focus on identifying robust EEG features that generalize across diverse driving conditions and investigating the impact of medications on EEG signals. Ultimately, this research supports the development of EEG-based indicators to assist drivers in maintaining alertness and reducing traffic casualties.

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
• 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