Related Experiment Video
Updated: Jun 26, 2026
06:34
A Single-Channel and Non-Invasive Wearable Brain-Computer Interface for Industry and Healthcare
Published on: July 7, 2023
Application of Filter Bank to Improve Fatigue Monitoring in Wearable EEG-Based Brain-Computer Interface
Timothy Jern Yu Tan1, Zhuo Zhang2, Kai Keng Ang2,3
1School of Chemistry, Chemical Engineering and Biotechnology (CCEB), Nanyang Technological University, 50 Nanyang Avenue, Singapore 639798, Singapore.
Neurosci
|June 25, 2026
Summary
A new filter bank approach significantly improves fatigue detection accuracy using wearable electroencephalography (EEG) brain-computer interfaces (BCIs). This method enhances monitoring for safety and performance by analyzing EEG signals more effectively.
Area of Science:
- Neuroscience
- Biomedical Engineering
- Signal Processing
Background:
- Fatigue monitoring is critical for preventing safety incidents by mitigating performance decline.
- Wearable brain-computer interfaces (BCIs) offer a promising avenue for real-time fatigue detection.
- Existing methods for fatigue detection using electroencephalography (EEG) signals can be further optimized.
Purpose of the Study:
- To propose and evaluate a novel filter bank-based approach for fatigue detection using EEG signals.
- To enhance the accuracy of fatigue detection in wearable EEG-based BCIs.
- To compare the proposed filter bank approach with a traditional broadband filter approach for fatigue detection.
Main Methods:
- Utilized a publicly available EEG dataset from 40 participants performing a Cognitive Vigilance Task (CVT).
- Decomposed EEG signals into delta, theta, alpha, beta, and gamma sub-bands using a filter bank for feature extraction.
- Trained two classification models: one with filter bank features and another with broadband filter features.
- Employed leave-one-subject-out cross-validation for performance evaluation.
Main Results:
- The filter bank approach achieved an accuracy of 86.4% ± 8.3%, outperforming the broadband filter approach (75.8% ± 10.4%).
- An overall accuracy increase of 10.6% was observed with the proposed filter bank method.
- The results highlight the superior performance of the filter bank approach in distinguishing fatigued from non-fatigued states.
Conclusions:
- The filter bank-based feature extraction method shows significant potential for improving fatigue detection in wearable EEG-based BCI systems.
- This approach offers enhanced accuracy and reliability for fatigue monitoring, contributing to improved safety and efficiency.
- Further research can explore the integration of this method into real-world applications for proactive fatigue management.
