Related Experiment Video
Updated: Nov 11, 2025
13:18
Conducting Concurrent Electroencephalography and Functional Near-Infrared Spectroscopy Recordings with a Flanker Task
Published on: May 24, 2020
7.9K
Toward practical driving fatigue detection using three frontal EEG channels: a proof-of-concept study
Xucheng Liu1,2, Gang Li3,4, Sujie Wang3
1Department of Electrical and Computer Engineering, Faculty of Science and Technology, University of Macau, Taipa, Macau.
Physiological Measurement
|March 29, 2021
Summary
This study introduces a practical driving fatigue detection method using non-hair-bearing (NHB) electroencephalogram (EEG) features. The NHB approach achieved high accuracy in identifying driver fatigue, paving the way for real-world applications.
Area of Science:
- Neuroscience
- Biomedical Engineering
- Human Factors Engineering
Background:
- Driving fatigue poses a significant safety risk, yet current detection methods lack practical applicability.
- Existing electroencephalogram (EEG) based fatigue detection often requires numerous channels and complex feature extraction.
Purpose of the Study:
- To develop a practical and efficient driving fatigue detection system using fewer EEG channels from non-hair-bearing (NHB) areas.
- To identify key EEG features from NHB regions that effectively discriminate between vigilant and fatigued states.
Main Methods:
- Recorded EEG data from 20 subjects during a 90-minute simulated driving task.
- Utilized a sliding-window approach to define individual vigilant and fatigued states.
- Extracted and classified features including power-spectrum density (PSD), functional connectivity (FC), and entropy from NHB EEG channels.
Main Results:
- The best classification performance was achieved using three EEG channel pairs from the NHB area.
- High within-subject detection rate (92.7%) and satisfactory cross-subject generalizability (77.13%) were obtained.
- Prominent features included PSD within the frontal NHB area and FC within/between frontal and parietal NHB areas.
Conclusions:
- The non-hair-bearing (NHB) EEG method offers a practical and effective approach for driving fatigue detection.
- This method enhances the efficiency and generalizability of fatigue detection, moving closer to real-world implementation.
