本研究提出一種基於AASM判讀規範並採用有限狀態機的睡眠階段自動辨識演算法。為了模仿技師在臨床上實際判讀的情形，本研究從腦電、眼動及肌電訊號中萃取三大類不同特徵：型態相關 (morphology related)、頻率相關 (frequency related)及時序相關 (temporal contextual information)特徵。型態相關特徵抓取訊號的波形、震幅與持續時長，頻率相關特徵則分析訊號中特定區間的頻率。時序相關特徵則探討不同特徵在單一圖譜中的位置及其對階段判讀的影響，上述類別特徵被用於建立分類樹以辨識睡眠階層。同時為了考慮睡眠階層判斷時需考慮時序的關係，本研究將睡眠多通道生理訊號中前一頁圖譜的資訊引入判斷，亦即利用有限狀態機結合分類樹進行睡眠階段辨識。針對每頁睡眠圖譜（30秒）判斷結果時，會根據前一頁圖譜的結果而有不同的決策邏輯。本研究使用國立成功大學附設醫院睡眠醫學中心提供的臨床睡眠資料NCKUH睡眠資料庫 (25位受測者)，以及Institute of Systems and Robotics at University of Coimbra所提供的ISRUC睡眠資料庫 (使用其中60位受測者)。各睡眠階層判斷平均準確率及Cohen’s kappa在NCKUH睡眠資料庫為75.8%及0.66；在ISRUC睡眠資料庫為71.92%及0.62。本研究所開發的演算法擷取出的型態類別、頻率類別、時序類別特徵同時也會標註在原多通道生理訊號上，使睡眠技師可連結並解釋多通道睡眠生理訊號中所出現的特徵及機器判讀的結果。

This research proposed a rule-based algorithm with finite state machine in view of the AASM scoring manual for sleep stage classification. To mimic how technicians score in real clinical situations, three kinds of features, morphology related feature, frequency related feature, and temporal contextual information, were extracted from EEG, EOG, and EMG. Morphology related feature captured the waveform, amplitude, and duration of the signal, while frequency related feature analyzed the frequency of the specific interval. As for temporal contextual information, it observed the location of the feature in an epoch and its influence on the stage classification. These three features will be utilized to construct a decision tree. Consider the information from previous epoch, a finite state machine was combined with decision tree to score the stages. For each epoch (30 seconds), the judgment would be different depend on the previous stage. The two databases were evaluated, including NCKUH database provided by the Sleep Center of National Cheng Kung University Hospital (25 subjects) and ISRUC database provided by the Institute of Systems and Robotics at University of Coimbra (60 subjects). The accuracy and Cohen’s kappa were 75.8% and 0.66 in NCKUH database and 71.92% and 0.62 in ISRUC database. The features extracted by the algorithm were marked on the raw signal and enable technicians to connect and interpret the results between the raw signal and the algorithm.

[1] Richard B. Berry et al., The AASM Manual for the Scoring of Sleep and Associated Events v2.6, 2020.
[2] Rechtschaffen A., A. Kales, eds. A manual of standardized terminology, techniques and scoring system for sleep stages of human subjects, CA: BI/BR, Los Angeles, 1968.
[3] F. de Carli et al., “A Method for the Automatic Detection of Arousals During Sleep,” Sleep, vol. 22, no. 5, pp. 561–572, 1999.
[4] T. Sugi, F. Kawana, and M. Nakamura, “Automatic EEG arousal detection for sleep apnea syndrome,” Biomedical Signal Processing and Control, vol. 4, no. 4, pp. 329–337, 2009.
[5] G. Bremer, J. R. Smith, and I. Karacan, “Automatic Detection of the K-Complex in Sleep Electroencephalograms,” IEEE Transactions on Biomedical Engineering, vol. BME-17, no. 4, pp. 314–323, 1970.
[6] A. Kam et al., “Detection of K-complexes in sleep EEG using CD-HMM. ” The 26th Annual International Conference of the IEEE Engineering in Medicine and Biology Society. Vol. 1, pp. 33-36, 2004.
[7] C. R. Patti et al., “K-complex detection based on pattern matched wavelets,” 2016 IEEE EMBS Conference on Biomedical Engineering and Sciences (IECBES), 2016.
[8] E. Huupponen et al., “Optimization of sigma amplitude threshold in sleep spindle detection,” Journal of Sleep Research, vol. 9, no. 4, pp. 327–334, 2000.
[9] E. Huupponen et al., “Development and comparison of four sleep spindle detection methods,” Artificial Intelligence in Medicine, vol. 40, no. 3, pp. 157–170, 2007.
[10] F. Duman et al., “Efficient sleep spindle detection algorithm with decision tree,” Expert Systems with Applications, vol. 36, no. 6, pp. 9980–9985, 2009.
[11] M. Massimini et al., “The Sleep Slow Oscillation as a Traveling Wave,” Journal of Neuroscience, vol. 24, no. 31, pp. 6862–6870, 2004.
[12] K. Takahashi and Y. Atsumi, “Precise Measurement of Individual Rapid Eye Movements in REM Sleep of Humans,” Sleep, vol. 20, no. 9, pp. 743–752, 1997.
[13] G. M. Hatzilabrou et al., “A comparison of conventional and matched filtering techniques for rapid eye movement detection of the newborn [electro-oculography],” IEEE Transactions on Biomedical Engineering, vol. 41, no. 10, pp. 990–995, 1994.
[14] B. D. Yetton et al., “Automatic detection of rapid eye movements (REMs): A machine learning approach,” Journal of Neuroscience Methods, vol. 259, pp. 72–82, 2016.
[15] R. J. McPartland, D. J. Kupfer, and F. Gordon Foster, “Rapid eye movement analyzer,” Electroencephalography and Clinical Neurophysiology, vol. 34, no. 3, pp. 317–320, 1973.
[16] E. Werth, D. J. Dijk, P. Achermann, and A. A. Borbely, “Dynamics of the sleep EEG after an early evening nap: experimental data and simulations,” American Journal of Physiology-Regulatory, Integrative and Comparative Physiology, vol. 271, no. 3, 1996.
[17] S. R. Ray et al., “Computer sleep stage scoring - an expert system approach,” International Journal of Bio-Medical Computing, vol. 19, no. 1, pp. 43–61, 1986.
[18] Haejeong Park et al., Kwangsuk Park, and Do-Un Jeong, “Hybrid neural-network and rule-based expert system for automatic sleep stage scoring,” Proceedings of the 22nd Annual International Conference of the IEEE Engineering in Medicine and Biology Society (Cat. No.00CH37143), vol. 2, pp. 1316 – 1319, 2000.
[19] Sheng-Fu Liang et al., “A rule-based automatic sleep staging method,” 2011 Annual International Conference of the IEEE Engineering in Medicine and Biology Society, vol. 205, pp. 169 – 176, 2011.
[20] S. A. Imtiaz and E. Rodriguez-Villegas, “Automatic sleep staging using state machine-controlled decision trees,” 2015 37th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), 2015.
[21] T. Sousa et al., “A two-step automatic sleep stage classification method with dubious range detection,” Computers in Biology and Medicine, vol. 59, Jan., pp. 42–53, 2015.
[22] P. Memar and F. Faradji, “A Novel Multi-Class EEG-Based Sleep Stage Classification System,” IEEE Transactions on Neural Systems and Rehabilitation Engineering, vol. 26, no. 1, Jan., pp. 84-95, 2018.
[23] X. Li et al., “HyCLASSS: A Hybrid Classifier for Automatic Sleep Stage Scoring,” IEEE Journal of Biomedical and Health Informatics, vol. 22, no. 2, pp. 375–385, 2018.
[24] G. Zhu, Y. Li, and P. Wen, “Analysis and Classification of Sleep Stages Based on Difference Visibility Graphs From a Single-Channel EEG Signal,” IEEE Journal of Biomedical and Health Informatics, vol. 18, no. 6, Feb., pp. 1813-1821, 2014.
[25] E. Alickovic and A. Subasi, “Ensemble SVM Method for Automatic Sleep Stage Classification,” IEEE Transactions on Instrumentation and Measurement, vol. 67, no. 6, Feb. 2018.
[26] H. Phan et al., “Joint Classification and Prediction CNN Framework for Automatic Sleep Stage Classification,” IEEE Transactions on Biomedical Engineering, vol. 66, no. 5, pp. 1285–1296, 2019.
[27] S. Chambon et al., “A Deep Learning Architecture for Temporal Sleep Stage Classification Using Multivariate and Multimodal Time Series,” IEEE Transactions on Neural Systems and Rehabilitation Engineering, vol. 26, no. 4, pp. 758–769, 2018.
[28] A. Supratak et al., “DeepSleepNet: A Model for Automatic Sleep Stage Scoring Based on Raw Single-Channel EEG,” IEEE Transactions on Neural Systems and Rehabilitation Engineering, vol. 25, no. 11, pp. 1998–2008, 2017.
[29] H. Dong et al.,” Mixed neural network approach for temporal sleep stage classification,” IEEE Transactions on Neural Systems and Rehabilitation Engineering, vol. 26, no. 2, pp. 324-333, 2017.
[30] S. Mousavi, F. Afghah, and U. R. Acharya, “SleepEEGNet: Automated sleep stage scoring with sequence to sequence deep learning approach,” PLOS ONE, 2019
[31] Andrew Morley, Lizzie Hill, Athanasios G. Kaditis, “10-20 system EEG Placement,” 2016. [Online]. Available: https://reurl.cc/Q9pAVb. [Accessed May 27, 2021].
[32] Wikipedia, “F1 score.” [Online]. Available: https://en.wikipedia.org/wiki/F1_score. [Accessed June 27, 2021].
[33] Boaz Shmueli, “Multi-Class Metrics Made Simple: the Kappa Score (aka Cohen’s Kappa Coefficient),” towardsdatascience.com, Dec 18, 2019. [Online]. Available: https://reurl.cc/nzmOpn. [Accessed June 27, 2021]
[34] W. R. Ruehland et al., “The new AASM criteria for scoring hypopneas: impact on the apnea hypopnea index,” Sleep, 2009.
[35] S. Khalighi, T. Sousa, J. M. Santos, and U. Nunes, “ISRUC-Sleep: A comprehensive public dataset for sleep researchers,” Computer Methods and Programs in Biomedicine, vol. 124, pp. 180–192, 2016.
[36] M. Buckland and F. Gey, “The relationship between Recall and Precision,” Journal of the American Society for Information Science, vol. 45, no. 1, pp. 12–19, 1994.
[37] B. Saletu, J. Grunberger, and P. Rajna, “Pharmaco-EEG profiles of antidepressants. Pharmacodynamic studies with fluvoxamine.,” British Journal of Clinical Pharmacology, vol. 15, no. S3, 1983.
[38] Lee-chiong, Teofilo, Sleep Medicine: Essentials and Review. Oxford University Press, USA, 2008, pp. 105.
[39] 劉勝義, “睡眠期的判讀,” in 睡眠醫學實務, Taiwan, 合記圖書出版社, 2011, ch. 13, pp. 197-234.