本研究旨在回應日益增長的工作負荷和疲勞對個人和社會帶來的挑戰。我們希望開發一種基於腦電圖(Electroencephalography, EEG)的疲勞偵測系統，以幫助人們更好地認識自己的疲勞狀態進而管理。過去已有許多與疲勞有關之分析，然而缺乏針對不同腦區在疲勞時之EEG特徵變化；儘管也有文獻透過視覺模擬量表(Visual Analogue Scales, VAS)評估疲勞程度，但並未將其與EEG數據進行分析；目前已有文獻用2-back任務有效檢測疲勞，但缺乏以靜息狀態作為基準進行個體校正，存在個體差異性；有些文獻能檢測疲勞，有些能評估疲勞程度，但尚未有文獻將兩者結合起來。
而為了解決以上問題，所以設計了用來讓受試者疲勞之持續性表現測驗(Continuous Performance Test, CPT)用來讓受試者疲勞，並且在CPT前與後都有靜息與用來檢測疲勞之2-back任務，也會透過VAS評估受試者疲勞程度。通過受試者參與實驗，收集了他們的EEG和VAS分數。我們運用深度學習和機器學習技術，提取並選擇最具代表性的EEG特徵。經過統計分析與VAS的分數相關性分析，我們發現在靜息狀態時疲勞與無疲勞theta、alpha頻帶之頻帶功率(Band Power, BP)在Frontal、Central、Right Temporal有正差異(p-value<0.05)，並且與VAS有正相關；在2-back任務下疲勞與無疲勞theta頻帶的BP以及?/?的功率比(Power Ratio, PR)有負差異(p-value<0.05)，也與VAS分數負相關。在靜息狀態的F4-T6這兩電極在beta頻帶的相位延遲指數(Phase Lag Index, PLI)在主體間模型（Inter-Subject Model）能夠有70%之疲勞檢測分類率，並且與VAS分數呈高度正相關（相關係數大於0.7），並且若只使用FP2-T6 theta-PLI、F8-TP8 beta-PLI這兩個特徵就能有100%分類率，並且與VAS具相關性，甚至使用深度學習可以只用前額三顆電極(FP1、FP2、Fz)達到90%分類率，檢測上更為方便。本研究還揭示了EEG和主觀疲勞感受之間的密切關聯。通過誘發疲勞之CPT下EEG與主觀疲勞程度的VAS評分進行相關性分析，也發現了前額三顆電極(FP1、FP2、Fz)之beta-BP、gamma-BP和碎形維度有高度相關性並且也有顯著性差異，能作為實時檢測之特徵。這些研究發現對於深入了解疲勞對腦波的影響以及疲勞監測和認知負載評估具有重要意義，未來可促使便攜式疲勞監測技術的開發和應用。

This study aims to address the challenges posed by the increasing workload and fatigue on individuals and society. We aim to develop a fatigue detection system based on electroencephalography (EEG) to assist individuals in better managing and understanding their fatigue levels. While previous research has explored various aspects of fatigue, there is a lack of research on EEG feature changes specific to fatigue in different brain regions. Although some studies have used visual analogue scales (VAS) to assess fatigue levels, they have not analyzed them in conjunction with EEG data. Existing literature has effectively detected fatigue using the 2-back task, but without a relaxation baseline for individual calibration, there are variations across individuals. While some studies can detect fatigue and others can assess fatigue levels, there is no integration of both aspects.
To address these issues, we designed a Continuous Performance Test (CPT) to induce fatigue in participants. Both relaxation tasks and 2-back tasks for fatigue detection were administered before and after the CPT, respectively. VAS was also used to evaluate participants' subjective fatigue levels. EEG data and VAS scores were collected from participants who took part in the experiment. We employed advanced deep learning and machine learning techniques to extract and select the most representative EEG features. Through statistical analysis and correlation analysis with VAS scores, we found significant differences (p-value < 0.05) in band power (BP) of the theta and alpha frequency bands between relaxed and non-fatigued states, positively correlated with VAS scores. During the 2-back task, we observed negative differences (p-value < 0.05) in the delta and theta band BP between fatigued and non-fatigued states, negatively correlated with VAS scores. In the inter-subject model, the phase lag index (PLI) of the beta band between the F4 and T6 electrodes during the relaxation state achieved a 70% classification rate for fatigue detection, highly correlated with VAS scores (correlation coefficient > 0.7). Using only the FP2-T6 theta-PLI and F8-TP8 beta-PLI features, a 100% classification rate was achieved. Additionally, with the use of deep learning, a 90% classification rate was achieved using only the three frontal electrodes (FP1, FP2, Fz), facilitating convenient fatigue detection.
This study also revealed the close association between EEG data and subjective fatigue perception. Through correlation analysis between EEG data during the CPT-induced fatigue and VAS scores, we found significant correlations and differences in beta-BP, gamma-BP, and fractal dimension of the three frontal electrodes (FP1, FP2, Fz), which can serve as real-time detection features. These findings are of great significance for gaining a deeper understanding of the impact of fatigue on brain waves and for fatigue monitoring and cognitive load assessment. They can potentially drive the development and application of portable fatigue monitoring technologies in the future.

[1] A. Gebel, A. Busch, C. Stelzel, T. Hortobágyi, and U. Granacher, “Effects of physical and mental fatigue on postural sway and cortical activity in healthy young adults,” Frontiers in Human Neuroscience, vol. 16, 2022.
[2] C. Anderson, and J. Bratman, “Translating thoughts into actions by finding patterns in brainwaves,” Computer Science, pp. 1-6, 2008.
[3] B. M. Tehrani, J. Wang, and D. Truax, “Assessment of mental fatigue using electroencephalography (EEG) and virtual reality (VR) for Construction Fall Hazard Prevention,” Engineering, Construction and Architectural Management, vol. 29, no. 9, pp. 3593–3616, 2021.
[4] B. T. Jap, S. Lal, P. Fischer, and E. Bekiaris, “Using EEG spectral components to assess algorithms for detecting fatigue,” Expert Systems with Applications, vol. 36, no. 2, pp. 2352–2359, 2009.
[5] P. Philip et al., “Fatigue, sleepiness, and performance in simulated versus real driving conditions,” Sleep, vol. 28, no. 12, pp. 1511–1516, 2005.
[6] S. Gündoğdu, A. Doğan, E. Gülbetekin, H. Çolak, and Ö. Polat, “Evaluation of the EEG signals and eye tracker data for working different N-back modes,” Traitement du Signal, vol. 36, no. 6, pp. 493–500, 2019.
[7] S. Pelegrina, M. T. Lechuga, J.A. García-Madruga, M. R. Elosúa, P. Macizo, M. Carreiras,L. J. Fuentes, and M. T. Bajo, “Normative data on the n-back task for children and young adolescents,” Front Psychol, vol. 6, no. 1544, pp. 1–11, 2015.
[8] L. Peng, J. Weng, and Y. Yang, “H. Wen, Impact of Light Environment on Driver’s Physiology and Psychology in Interior Zone of Long Tunnel,” Frontiers in Public Health, vol. 10, 2022.
[9] F. Noé, B. Hachard, H. Ceyte, N. Bru, and T. Paillard, “Relationship between the level of mental fatigue induced by a prolonged cognitive task and the degree of Balance Disturbance,” Experimental Brain Research, vol. 239, no. 7, pp. 2273–2283, 2021.
[10] A. A. Bennett, A. B. Bakker, and J. G. Field, “Recovery from work-related effort: A meta-analysis,” Journal of Organizational Behavior, vol. 39, no. 3, pp. 262–275, 2017.
[11] P. C. Winwood, A. H. Winefield, and K. Lushington, “Work-related fatigue and recovery: The contribution of age, domestic responsibilities and Shiftwork,” Journal of Advanced Nursing, vol. 56, no. 4, pp. 438–449, 2006.
[12] T. F. Meijman, “Mental fatigue and the efficiency of information processing in relation to Work Times,” International Journal of Industrial Ergonomics, vol. 20, no. 1, pp. 31–38, 1997.
[13] T. Åkerstedt, “Mental fatigue, work and sleep,” Journal of Psychosomatic Research, 2004.
[14] National Safety Council (NSC). “What is Fatigue Costing Your Company?.” https://www.nsc.org/workplace/safety-topics/fatigue/what-is-fatigue-costing-your-company (accessed July. 3, 2023).
[15] O. Wilhelm, A. Hildebrandt, and K. Oberauer, “What is working memory capacity, and how can we measure it?,” Front Psychol, vol. 433, no. 4, pp. 1–22, 2013.
[16] M. Tanaka, Y. Shigihara, A. Ishii, M. Funakura, E. Kanai, and Y. Watanabe, “Effect of mental fatigue on the central nervous system: An electroencephalography study,” Behav Brain Funct vol. 8, no. 48, pp. 1–8, 2012.
[17] J. Lim and D. F. Dinges, “A meta-analysis of the impact of short-term sleep deprivation on cognitive variables.,” Psychological Bulletin, vol. 136, no. 3, pp. 375–389, 2010.
[18] Y. Peng, C. Li, Q. Chen, Y. Zhu and L. Sun, “Functional Connectivity Analysis and Detection of Mental Fatigue Induced by Different Tasks Using Functional Near-Infrared Spectroscopy,” Frontiers in Neuroscience, vol. 15, 2022.
[19] M. A. S. Boksem, T. F. Meijman, and M. M. Lorist, “Mental fatigue, motivation and action monitoring,” Biological Psychology, vol. 72, no. 2, pp. 123–132, 2006.
[20] M. Jagannath and V. Balasubramanian, “Assessment of early onset of driver fatigue using multimodal fatigue measures in a static simulator,” Applied Ergonomics, vol. 45, no. 4, pp. 1140–1147, 2014.
[21] M. E. Smith, A. Gevins, H. Brown, A. Karnik, and R. Du, “Monitoring task loading with multivariate EEG measures during complex forms of human-computer interaction,” Human Factors: The Journal of the Human Factors and Ergonomics Society, vol. 43, no. 3, pp. 366–380, 2001.
[22] S. Kar, M. Bhagat, and A. Routray, “EEG signal analysis for the assessment and quantification of driver’s fatigue,” Transportation Research Part F: Traffic Psychology and Behaviour, vol. 13, no. 5, pp. 297–306, 2010.
[23] C. Zhao, M. Zhao, J. Liu, and C. Zheng, “Electroencephalogram and electrocardiograph assessment of mental fatigue in a driving simulator,” Accident Analysis & Prevention, vol. 45, pp. 83–90, 2012.
[24] H. J. Eoh, M. K. Chung, and S.-H. Kim, “Electroencephalographic study of drowsiness in simulated driving with sleep deprivation,” International Journal of Industrial Ergonomics, vol. 35, no. 4, pp. 307–320, 2005.
[25] A. Ekim, O. Aydemir and M. Demir, "Classification of Cognitive Fatigue with EEG Signals," 2020 Medical Technologies Congress (TIPTEKNO), 2020.
[26] L. Chen, Y. Zhao, J. Zhang, and J. Zou, “Automatic detection of alertness/drowsiness from physiological signals using wavelet-based nonlinear features and Machine Learning,” Expert Systems with Applications, vol. 42, no. 21, pp. 7344–7355, 2015.
[27] J. Min, P. Wang, and J. Hu, “Driver fatigue detection through multiple entropy fusion analysis in an EEG-based system,” PLOS ONE, vol. 12, no. 12, 2017.
[28] P. Qi et al., “EEG Functional Connectivity Predicts Individual Behavioural Impairment During Mental Fatigue,” IEEE Transactions on Neural Systems and Rehabilitation Engineering, vol. 28, no. 9, pp. 2080-2089, 2020.
[29] J.-P. Liu, C. Zhang and C.-X. Zheng, “Estimation of the cortical functional connectivity by directed transfer function during mental fatigue”, Applied Ergonomics., vol. 42, no. 1, pp. 114-121, 2010.
[30] M. S. Clayton, N. Yeung and R. Cohen Kadosh, “The roles of cortical oscillations in sustained attention”, Trends Cognit. Sci., vol. 19, no. 4, pp. 188-195, 2015.
[31] A. K. Kumawat, “Inter and intra-hemispheric resting EEG coherence in schizophrenia patients,” Journal of Medical Science And clinical Research, vol. 08, no. 01, 2020.
[32] Y. Chen, W. Fang, B. Guo, and H. Bao, “Fatigue-related effects in the process of task interruption on working memory,” Frontiers in Human Neuroscience, vol. 15, 2021.
[33] J. H. Kim, Y. Cho, Y. -A. Suh and M. -S. Yim, “Development of an Information Security-Enforced EEG-Based Nuclear Operators’ Fitness for Duty Classification System,” IEEE Access, vol. 9, pp. 72535-72546, 2021.
[34] V. J. Lawhern, A. J. Solon, N. R. Waytowich, S. M. Gordon, C. P. Hung and B. J. Lance, "EEGNet: A compact convolutional neural network for EEG-based brain–computer interfaces", Journal of Neural Engineering, vol. 15, no. 5, 2018.
[35] A. Asif, M. Majid, and S. M. Anwar, “Human stress classification using EEG signals in response to music tracks,” Computers in Biology and Medicine, vol. 107, pp. 182-196, 2019.
[36] Y. Zhang, Q. Wang, Z. Y. Chin, and K. K. Ang, “Investigating different stress-relief methods using Electroencephalogram (EEG),” 2020 42nd Annual International Conference of the IEEE Engineering in Medicine & Biology Society (EMBC), 2020.
[37] T. Mullen et al., “Real-time modeling and 3D visualization of source dynamics and connectivity using wearable EEG,” 2013 35th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), 2013.
[38] C. Liu, W. Gaetz, and H. Zhu, “Estimation of time-varying coherence and its application in understanding brain functional connectivity,” EURASIP Journal on Advances in Signal Processing, vol. 2010.
[39] C. J. Stam, G. Nolte, and A. Daffertshofer, “Phase lag index: Assessment of functional connectivity from multi-channel EEG and Meg with diminished bias from common sources,” Human Brain Mapping, vol. 28, no. 11, pp. 1178–1193, 2007.
[40] S. Mika, G. Ratsch, J. Weston, B. Scholkopf and K. R. Mullers, "Fisher discriminant analysis with kernels," Neural Networks for Signal Processing IX: Proceedings of the 1999 IEEE Signal Processing Society Workshop (Cat. No.98TH8468), Madison, WI, USA, 1999.
[41] L.-W. Ko, O. Komarov, W.-K. Lai, W.-G. Liang, and T.-P. Jung, “Eyeblink recognition improves fatigue prediction from single-channel forehead EEG in a realistic sustained attention task,” Journal of Neural Engineering, vol. 17, no. 3, p. 036015, 2020.