Classifying Driving Fatigue Based on Combined Entropy Measure Using EEG Signals
DOI: 10.14257/ijca.2016.9.3.30
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 addresses the critical safety issue of driver fatigue, a major contributor to road accidents, by developing a method to automatically classify alert and drowsy states using electroencephalogram (EEG) signals. While previous research has utilized frequency-domain analysis or linear methods, this work focuses on nonlinear complexity measures to better capture the irregularity and randomness of brain activity associated with fatigue. The authors propose combining Approximate Entropy (AE) and Sample Entropy (SE) as feature extraction metrics, followed by classification using a Support Vector Machine (SVM), aiming to create a robust detection system for real-time monitoring. The experimental design involved sixty graduate students who participated in a driving simulation study. Participants first completed a 15-minute alert driving session with high visual stimuli to establish a baseline. This was followed by a monotonous driving session lasting 2–3 hours, designed to induce drowsiness. EEG signals were recorded from 14 channels using the international 10-20 system, along with eye movement data to assist in labeling. Data preprocessing included artifact removal and segmentation into 10-second epochs. An independent psychophysiologist labeled epochs as "alert" or "drowsy" based on eye blink patterns and dominant EEG frequency components. From the collected data, 4,000 alert and 4,000 drowsy epochs were selected for analysis. AE and SE were calculated for each epoch to quantify signal complexity, and these combined features were fed into an SVM classifier trained and tested using 10-fold cross-validation. The results demonstrated that both AE and SE values significantly decreased as fatigue levels increased, indicating reduced complexity in EEG signals during drowsiness. Statistical analysis revealed significant differences between alert and drowsy states primarily at parietal (P3, P4) and occipital (Oz) electrodes. When using the combined AE and SE features from these specific channels, the SVM classifier achieved high performance. The highest averaged classification accuracy was obtained at the P3 electrode (91.28%), with a sensitivity of 93.67% for detecting drowsiness and a specificity of 88.89% for identifying alert states. The study found that while single entropy measures showed limited discriminatory power individually, their combination significantly enhanced classification accuracy. The significance of this work lies in demonstrating that nonlinear entropy measures, particularly when combined, are effective indicators of driver fatigue. The findings suggest that EEG signals from parietal and occipital regions are highly sensitive to changes in alertness. The proposed SVM-based classification method offers a reliable, automated approach for detecting drowsiness, with potential applications in onboard safety systems for professional drivers. The authors conclude that this nonlinear analysis provides deeper insights into brain dynamics than traditional linear methods, though future research should focus on optimizing channel selection for ergonomic feasibility and expanding the subject pool to include diverse demographics.
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.
| Stage | Outcome | Tool | Model | Prompt | Attempts | Completed |
|---|---|---|---|---|---|---|
| discover | success | Crossref | — | — | 1 | 2026-08-09 |
| archive | success | canonical_url | — | — | 1 | 2026-08-09 |
| extract | success | cached | — | — | 3 | 2026-08-10 |
| clean | success | clean | — | — | 1 | 2026-08-09 |
| chunk | success | chunk | — | — | 1 | 2026-08-09 |
| embed | success | embed | Qwen/Qwen3-Embedding-8B | — | 1 | 2026-08-09 |
| promote | success | — | — | — | 1 | 2026-08-09 |
| summarize | success | llm | qwen3.6-27b-nvidia | summ-v5 | 2 | 2026-08-10 |
| tag | success | vector_similarity | — | — | 10 | 2026-08-11 |
| verify | partial | — | — | — | 1 | 2026-08-10 |
Summary generated by qwen3.6-27b-nvidia on 2026-08-10; verification: verified_with_issues.
Topics
Ranked by relevance to this paper. Hover a topic for its definition.
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
- Methodological Resource: validation psychometrics