基于深度学习的脑电信号自动睡眠分期研究进展
Research Progress of EEG Automatic Sleep Staging Based on Deep Learning
DOI: 10.12677/AAM.2023.121004, PDF, HTML, XML, 下载: 461  浏览: 1,769  科研立项经费支持
作者: 许 哲, 章浩伟*, 刘 颖:上海理工大学健康科学与工程学院,上海
关键词: 睡眠脑电信号深度学习自动分期Sleep Eeg Signal Deep Learning Automatic Staging
摘要: 睡眠是人类不可或缺的生理活动,准确地睡眠分期是诊断睡眠疾病的前提。当前,基于深度学习的脑电信号自动睡眠分期正成为研究的热点,虽然相关研究取得很多进展,但距离临床应用还有一定距离。本文就该领域展开综述,详细介绍了近年来基于深度学习的脑电信号自动睡眠分期方法,综合论述目前主流神经网络在自动睡眠分期领域的研究现状及进展,分析归纳了不同模型网络的潜力优势及未来发展方向,以促进深度学习技术在基于脑电信号的自动分期研究更深入发展。
Abstract: Sleep is an indispensable physiological activity of human beings. Accurate sleep staging is the premise of diagnosing sleep diseases. At present, EEG automatic sleep staging based on deep learn-ing is becoming a hot research topic. Although related researches have made a lot of progress, there is still a long way to go before clinical application. This paper reviews this field, introduces in detail the EEG automatic sleep staging methods based on deep learning in recent years, comprehensively discusses the current research status and progress of mainstream neural networks in the field of automatic sleep staging, analyzes and summarizes the potential advantages and future develop-ment direction of different model networks. In order to promote the deep learning technology in the automatic staging based on EEG further development.
文章引用:许哲, 章浩伟, 刘颖. 基于深度学习的脑电信号自动睡眠分期研究进展[J]. 应用数学进展, 2023, 12(1): 21-28. https://doi.org/10.12677/AAM.2023.121004

1. 引言

睡眠在人类的健康中起着至关重要的作用 [1]。我们生命的三分之一时间都处于睡眠状态,良好的睡眠质量、充足的睡眠时间和完整的睡眠结构有利于调节机体免疫功能,维持各系统功能处于稳定状态,对于保持一个人的精神和身体健康至关重要 [2]。随着现代生活压力的逐渐增大,越来越多的人面临睡眠问题的困扰,甚至有些患上了睡眠障碍疾病,许多的心血管疾病与精神疾病也都与睡眠问题相关 [3]。人们现在越来越意识到睡眠在保护我们的身心健康方面的重要作用,越来越重视睡眠问题。临床医生通过不同的睡眠周期、每个睡眠阶段的时长、睡眠潜伏期等表明的神经生理过程来判断睡眠障碍、阻塞性睡眠呼吸暂停等疾病 [4]。作为研究睡眠问题的基础,完成睡眠质量评估的前提,只有对睡眠进行正确的分期,才能进一步研究睡眠问题。在临床诊断上,睡眠分期主要由专家进行手工分期 [5],但在分期过程中个人主观性和不可预测性较强,且整晚睡眠的时间较长,导致手工睡眠分期效率低且错误率高 [6]。因此,怎样利用计算机技术使睡眠分期更加精确、高效,是一个重要的研究内容。

近年来,随着计算机技术的不断发展,深度学习技术在计算机视觉、语音识别、自然语言处理等方面得到了极为广泛的应用 [7] [8] [9] [10] [11]。在自动睡眠分期研究方面,各种不同的神经网络模型也取得了较好的睡眠分期准确率 [12] [13]。目前,基于脑电信号的自动睡眠研究是睡眠研究中最为活跃的部分,一个重要原因是睡眠时期脑电不受主观思维的影响,更能反映出对睡眠的控制特征。该研究课题也因为其理论背景和应用价值,引起了来自生物、医学、计算机等众多领域研究者的广泛关注。

本文针对基于深度学习的脑电信号自动睡眠分期研究,首先介绍了脑电信号与睡眠分期基础知识,然后介绍了深度学习技术在脑电信号自动睡眠分期研究方面的最新研究概述和技术见解。最后归纳了目前存在的问题并对其发展趋势进行展望。

2. 脑电信号与睡眠分期基础知识

脑电(electroencephalogram, EEG)信号是大脑内同步神经元活动产生的微伏级电信号,其波幅一般不会超过200 uV,包含生理和行为两方面的特征信息 [14]。在临床应用中,通常使用多导睡眠仪(polysomnography, PSG)收集和记录患者的整晚脑电信号。PSG的电极放置采用国际标准的10~20系统电极放置法 [15],即首先确定鼻根到枕骨部之间、两只耳朵左右方向之间的连线,两条直线的交点是Cz电极,从鼻子根部往后量大概10%的地方是FPz电极,从FPz再向后量每一个20%都放置一个电极,依次定义位置Fz、Cz、Pz以及Oz。大多数研究者选择FPz-Cz通道进行实验,因为该通道包含的脑电信息更为丰富,是脑电记录中的典型通道。

睡眠分期是从PSG监测到的生理信号中提取睡眠周期信息的过程。1968年,Rechtschaffen和Kales最早制定和出版了相对统一和标准的睡眠分期标准,即R&K判读规则 [16]。它将睡眠分成了不同的阶段:清醒期(wake, W)、非快速眼球运动(non rapid eye movement, NREM)、快速眼球运动(rapid eye movement, REM)阶段。其中NREM又被分为I~IV期,分别代表睡眠由浅入深过程的不同阶段。目前,国际上普遍采用2007年美国睡眠医学会(The American Academy of Sleep Medical, AASM)修正过的R&K睡眠分期标准,将睡眠过程修改为清醒期(W)、浅睡期(N1、N2)、深睡期(N3)、和快速眼球运动期(REM) [17]。R&K标准与AASM标准的关系如图1所示。

Figure 1. Relationship between R&K standard and AASM standard

图1. R&K标准与AASM标准关系图

3. 基于深度学习的脑电信号自动睡眠分期研究

3.1. 公开数据库

常用于自动睡眠阶段分期的公开脑电数据库有六个。其中五个数据库可从PhysioNet上免费下载 [18],分别是睡眠集–欧洲数据格式(sleep-european data format, Sleep-EDF) [19]、扩充的睡眠数据集(sleep-EDF database expanded, SLEEP-EDFx)、睡眠心脏健康研究(the sleep heart health study, SHHS) [20]、麻省理工学院贝斯以色列医院数据库(MIT-BIH) [21] 和ISRUC-Sleep数据集 [22]。蒙特利尔睡眠研究数据(the montrealarchive of sleep studie, MASS)需要获得许可下载 [23]。表1列出了数据集的获取地址及简介。

3.2. 神经网络分类器

深度学习技术可以从训练数据中自动提取数据特征用于分类,并能随着数据量的增加提高模型的性能,使其分类结果更加精确、高效,在自动睡眠分期研究方面取得了较好结果。常用的基于深度学习自动睡眠方法包括3类:1) 卷积神经网络模型(convolu-tionalneural network, CNN);2) 循环神经网络模型(recurrent neural network, RNN);3) 混合神经网络模型(hybrid neural networks)。

Table 1. Open dataset and introduction

表1. 公开数据集及简介

3.2.1. 卷积神经网络

CNN是至少具有一个卷积层的前馈神经网络,主要有卷积层、池化层、全连接层组成,被广泛应用于图像处理领域 [24] [25]。其中,卷积层负责提取图像中的局部特征;池化层用来大幅降低参数量级;全连接层类似传统神经网络的部分,用来输出想要的结果 [26]。相比其他神经网络模型,卷积网络的特征提取能力更强,而且相对易于训练。CNN也是最先应用于脑电信号领域的深度学习模型 [27]。

由于脑电信号在采集过程中会使用多个通道,所以很多研究将多通道脑电信号映射为二维(two-dimensional, 2D)或三维(three-dimensional, 3D)图像,以便于采用CNN进行模型构建。Tsinalis等 [28] 构建具有两对卷积层和池化层,两个全连接层的CNN模型,结合Softmax函数,实现了正常睡眠阶段的自动分期。Zhu等 [29] 将注意力机制引入了CNN网络中共同执行自动睡眠分期,模型使用Sleep-EDF数据集实现了93.7%的总体准确性。然而,当对扩展的Sleep-EDFx数据库中的EEG信号进行测试时,同样的模型只获得了82.8%的准确性。Zhang等 [30] 构建具有五层CNN结构的神经网络模型,使用临床收集的数据集,实现了96%的总体准确性。但是,这种准确性是通过使用私人临床数据集实现的,当他们使用Sleep-EDF数据集评估模型的性能时,总体准确率达到了86.4%,这低于Zhu等的模型。为了避免因更有效的捕捉特征而增加网络深度引起的梯度消失问题,Zhu等和Cui等 [31] 选择使用层数较少的CNN模型,通过使用注意力机制和多尺度熵中的细粒度段来增加模型特征提取能力,从而获得较高的睡眠阶段分类性能。

3.2.2. 循环神经网络

CNN虽然能充分挖掘数据的时频域特征,但无法提取数据间的时序特征,而EEG信号是具有高度随机性的非线性时间序列数据 [32],不仅包含时频域特征,还包含时序特征。RNN正是用于处理时间序列数据的网络,其在自然语言处理、机器翻译等领域获得了广泛应用 [33] [34] [35]。

为了能够充分利用脑电信号的时序信息,一些研究采用 RNN来构建自动睡眠分期模型。Hsu等 [36] 采用Elman网络结构,构建4层RNN网络模型,成功地对睡眠的各个阶段进行了分类,实现了模型最佳性能,总体准确率为87.2%。然而由于传统RNN模型容易产生梯度消失等问题,无法学习长期依赖关系,且训练效率低下,所以研究者们大多使用RNN的变体网络,如长短期记忆(longshort-termmemory, LSTM)网络和双向长短期记忆网络(bi-directional long short-term memory, BiLSTM)等。这些网络不仅能有效挖掘脑电数据中的时序特征,还解决了梯度消失问题。如Michielli等人 [32] 提出了一个具有2个LSTM单元的级联RNN网络,获得了类似的86.7%的精度。BiLSTM是LSTM的一种变体,相比于单向LSTM而言,BiLSTM能够同时利用过去时刻和未来时刻的信息,比单向LSTM获得更准确的预测效果。这在自动睡眠分期过程中是合理的,因为专家在睡眠分期过程中,不仅要考虑这一帧的数据信息,还要考虑前后帧对其的影响。You等 [37] 将BiLSTM应用于所提出的模型,将模型的整体准确率提高了约1%,达到了81.6%,所提出的模型的参数测量值仅为0.31 MB,是DeepSleepNet [38] 的5%,但其性能与DeepSleepNet相似。Fu等 [39] 整合BiLSTM网络,使用单一EEG通道获得了总体83.78%的分类精度。

3.2.3. 混合神经网络

为了充分利用卷积神经网络在特征选择、提取方面的良好性能,循环神经网络处理具有时序信息数据时的独特优势,进一步提高自动睡眠分期模型的性能,很多研究者提出了将两者相结合的深度模型,这也是最近研究的热点工作 [40] [41] [42]。Amelia等 [43] 构建了包含三个卷积层和两个LSTM层的深度学习模型,对低质量双通道脑电数据的验证准确率为74% (±10%),在黄金标准PSG上实现了77% (±10%)的验证准确率。Zhao等 [44] 搭建一维CNN-LSTM模型,使用Fpz-Cz通道获得了93.47%的分期准确率。为了在特征提取中充分考虑局部特征,并且区分关键和非关键局部特征的重要性,Tingting等 [45] 在由CNN和BiLSTM组成的CAttSleepNet模型中加入了注意力机制,加强EEG信号局部和全局上下文相关性特征,在Sleep-EDF和Sleep-EDFx数据集上均获得了优于其它模型的实验结果。

4. 基于深度学习的脑电信号自动睡眠研究存在的问题

综上所述,深度学习技术打破了传统机器学习难以克服的问题,在基于脑电信号的自动睡眠分期研究中已经有了很多成功的运用,但是仍未达到与临床专家手工划分一致的水平,将其应用于临床应用仍面临一系列问题。

在模型的结构选择方面,当前主要流行的网络结构并不是专门为分析脑电数据而设计的。研究者为了适应网络需要对输入的脑电信号进行额外的处理,可能会导致在模型训练过程中无法充分利用脑电信号的原始信息,因此依据不同分类器的独特优势,选择或设计单一或混合的神经网络分类器依然是提升脑电自动睡眠分期模型性能的关键。

深度学习模型需要大量的数据进行训练才能表现出较好的性能,而睡眠脑电数据在采集过程中要求受试者进行长时间的睡眠,这无疑增加了获取数据的难度。当网络结构非常深、非常复杂时,数据量较少会极大影响模型的性能。而可供自动睡眠分期模型训练的临床脑电数据较少,依然制约着深度学习技术在自动睡眠分期领域的发展。

5. 总结与展望

本文根据基于深度学习的脑电信号自动睡眠分期研究,介绍了脑电信号及睡眠分期相关基础知识,以及实验常用的公开数据库,对比分析介绍了最新研究方法。未来深度学习在脑电自动睡眠分期的研究可能会着重朝以下方向发展:通过改进优化网络结构使其适用于脑电信号,以增强模型对于输入数据的契合度,混合神经网络即是此方向发展的实例;通过精简模型的参数,充分利用现有脑电数据,实现基于小样本的自动睡眠分期;通过研发更为便捷快速的脑电信号采集技术,扩大样本体量,这是从根本上解决脑电数据量问题方法。

深度学习技术在基于脑电的自动睡眠分期研究领域已经成为当前发展趋势。未来随着深度学习技术自身的进步,将会有更多优秀模型及方法出现,实现高效、精准的自动睡眠分期。

基金项目

上海介入医疗器械工程技术研究中心(18DZ2250900);上海理工大学医工交叉项目(1021308424)。

NOTES

*通讯作者。

参考文献

[1] Czeisler, C.A. (2015) Duration, Timing and Quality of Sleep Are Each Vital for Health, Performance and Safety. Sleep Health: Journal of the National Sleep Foundation, 1, 5-8.
https://doi.org/10.1016/j.sleh.2014.12.008
[2] Phan, H., Andreotti, F., Cooray, N., Chén, O.Y. and De Vos, M. (2018) Joint Classification and Prediction CNN Framework for Automatic Sleep Stage Classification. IEEE Transactions on Biomedical Engineering, 66, 1285-1296.
https://doi.org/10.1109/TBME.2018.2872652
[3] Phan, H., Mikkelsen, K., Chén, O.Y., et al. (2022) SleepTrans-former: Automatic Sleep Staging with Interpretability and Uncertainty Quantification. IEEE Transactions on Biomedical Engineering, 69, 2456-2467.
https://doi.org/10.1109/TBME.2022.3147187
[4] Phan, H. and Mikkelsen, K. (2022) Automatic Sleep Staging of EEG Signals: Recent Development, Challenges, and Future Directions. Physiological Measurement, 43, Article ID: 04TR01.
https://doi.org/10.1088/1361-6579/ac6049
[5] Mousavi, S., Afghah, F. and Acharya, U.R. (2019) SleepEEGNet: Automated Sleep Stage Scoring with Sequence to Sequence Deep Learning Approach. PLOS ONE, 14, e0216456.
https://doi.org/10.1371/journal.pone.0216456
[6] Clements-Cortes, A and Kim, C.T. (2017) Music for a Better Night’s Sleep. Canadian Music Educator, 58, 33-36.
[7] Ma, N., Wu, Z., Cheung, Y. M., Guo, Y., Gao, Y., Li, J. and Jiang, B. (2022) A Survey of Human Action Recognition and Posture Prediction. Tsinghua Science and Tech-nology, 27, 973-1001.
https://doi.org/10.26599/TST.2021.9010068
[8] 朱方圆, 马志强, 陈艳, 张晓旭, 王洪彬, 宝财吉拉呼. 语音识别中说话人自适应方法研究综述[J]. 计算机科学与探索, 2021, 15(12): 2241-2255.
[9] Shankar, V. and Parsa-na, S. (2022) An Overview and Empirical Comparison of Natural Language Processing (NLP) Models and an Introduc-tion to and Empirical Application of Autoencoder Models in Marketing. Journal of the Academy of Marketing Science, 50, 1324-1350.
https://doi.org/10.1007/s11747-022-00840-3
[10] Kamath, S., Karibasappa, K.G., Reddy, A., Kallur, A.M., Priyanka, B.B. and Bhagya, B.P. (2021) Improving the Relation Classification Using Convolutional Neural Net-work. IOP Conference Series: Materials Science and Engineering, 1187, Article ID: 012004.
https://doi.org/10.1088/1757-899X/1187/1/012004
[11] Li, Q., Chen, Y. and Zeng, Y. (2022) Transformer with Transfer CNN for Remote-Sensing-Image Object Detection. Remote Sensing, 14, Article No. 984.
https://doi.org/10.3390/rs14040984
[12] Ek, A. and Bma, A. (2021) Automatic Sleep Stage Classification Using Temporal Convolutional Neural Network and New Data Augmentation Technique from Raw Single-Channel EEG. Computer Methods and Programs in Biomedicine, 204, Article ID: 106063.
[13] Phan, H., Andreotti, F., Cooray, N., Chén, O.Y. and De Vos, M. (2018) Automatic Sleep Stage Classification Using Single-Channel EEG: Learning Sequen-tial Features with Attention-Based Recurrent Neural Networks. 2018 40th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, Honolulu, 18-21 July 2018, 1452-1455.
https://doi.org/10.1109/EMBC.2018.8512480
[14] Casson, A.J., Yates, D.C., Smith, S.J.M., et al. (2010) Weara-ble Electroencephalography. IEEE Engineering in Medicine and Biology Magazine, 29, 44-56.
https://doi.org/10.1109/MEMB.2010.936545
[15] Jasper, H.H. (1958) The 10-20 Electrode System of the Interna-tional Federation. Electroencephalography and Clinical Neurophysiology, 10, 370-375.
https://doi.org/10.1016/0013-4694(58)90053-1
[16] Wolpert, E.A. (1969) A Manual of Standardized Terminology, Techniques and Scoring System for Sleep Stages of Human Subjects. Electroencephalography & Clinical Neurophysi-ology, 26, 644-644.
https://doi.org/10.1016/0013-4694(69)90021-2
[17] Iber, C., Ancoli-Israel, S., Chesson, A.L., et al. (2015) The AASM Manual for the Scoring of Sleep and Associated Events: Rules, Terminology and Technical Specifications. American Academy of Sleep Medicine, Westchester IL.
[18] Goldberger, A.L., Amaral, L.A., Glass, L., et al. (2000) PhysioBank, PhysioToolkit, and PhysioNet: Components of a New Research Resource for Complex Physiologic Signals. Circulation, 101, e215-e220.
https://doi.org/10.1161/01.CIR.101.23.e215
[19] Kemp, B., Zwinderman, A.H., Tuk, B., et al. (2000) Analysis of a Sleep-Dependent Neuronal Feedback Loop: The Slow-Wave Microcontinuity of the EEG. IEEE Transactions on Bio-medical Engineering, 47, 1185-1194.
https://doi.org/10.1109/10.867928
[20] Zhang, G.-Q., Cui, L., Mueller, R., et al. (2018) The National Sleep Re-search Resource: Towards a Sleep Data Commons. Journal of the American Medical Informatics Association, 25, 1351-1358.
https://doi.org/10.1093/jamia/ocy064
[21] Ichimaru, Y and Moody, G.B. (1999) Development of the Polysomnographic Database on CD-ROM. Psychiatry and Clinical Neurosciences, 53, 175-177.
https://doi.org/10.1046/j.1440-1819.1999.00527.x
[22] Khalighi, S., Sousa, T., Santos, J.M. and Nunes, U. (2016) ISRUC-Sleep: A Comprehensive Public Dataset for Sleep Researchers. Computer Methods and Programs in Biomedi-cine, 124, 180-192.
https://doi.org/10.1016/j.cmpb.2015.10.013
[23] O’Reilly, C., Gosselin, N., Carrier, J. and Nielsen, T. (2014) Montreal Archive of Sleep Studies: An Open-Access Resource for Instrument Benchmarking and Exploratory Research. Journal of Sleep Research, 23, 628-635.
https://doi.org/10.1111/jsr.12169
[24] Ye, F., Xu, S., Wang, T., Wang, Z. and Ren, T. (2021) Application of CNN Algorithm Based on Chaotic Recursive Diagonal Model in Medical Image Processing. Computational Intelligence and Neuroscience, 2021, Article ID: 6168562.
https://doi.org/10.1155/2021/6168562
[25] Hu, X., Shi, W., Zhou, Y., Tang, H. and Duan, S. (2022) Quantized and Adaptive Memristor Based CNN (QA-mCNN) for Image Processing. Sci-ence China Information Sciences, 65, Article No. 119104.
https://doi.org/10.1007/s11432-020-3031-9
[26] LeCun, Y. and Bengio, Y. (1995) Convolutional Networks for Images, Speech, and Time Series. In: Arbib, M.A., Ed., The Handbook of Brain Theory and Neural Networks, MIT Press, Cambridge, 255-258.
[27] Cecotti, H. and Graser, A. (2010) Convolutional Neural Networks for P300 Detec-tion with Application to Brain-Com- puter Interfaces. IEEE Transactions on Pattern Analysis and Machine Intelligence, 233, 433-445.
https://doi.org/10.1109/TPAMI.2010.125
[28] Tsinalis, O., Matthews, P.M., Guo, Y. and Zafeiriou, S. (2016) Automatic Sleep Stage Scoring with Single-Channel EEG Using Convolutional Neural Networks. ArXiv Preprint ArXiv: 1610. 01683.
[29] Zhu, T., Luo, W. and Yu, F. (2020) Convolution- and Attention-Based Neural Network for Auto-mated Sleep Stage Classification. International Journal of Environmental Research and Public Health, 17, Article No. 4152.
https://doi.org/10.3390/ijerph17114152
[30] Zhang, X., Xu, M., Li, Y., Su, M., et al. (2020) Automated Mul-ti-Model Deep Neural Network for Sleep Stage Scoring with Unfiltered Clinical Data. Sleep and Breathing, 24, 581-590.
https://doi.org/10.1007/s11325-019-02008-w
[31] Cui, Z., Zheng, X., Shao, X. and Cui, L. (2018) Automatic Sleep Stage Classification Based on Convolutional Neural Network and Fine-Grained Segments. Complexity, 2018, Arti-cle ID: 9248410.
https://doi.org/10.1155/2018/9248410
[32] Michielli, N., Acharya, U.R. and Molinari, F. (2019) Cascaded LSTM Recurrent Neural Network for Automated Sleep Stage Classification Using Single-Channel EEG Sig-nals. Computers in Biology and Medicine, 106, 71-81.
https://doi.org/10.1016/j.compbiomed.2019.01.013
[33] Lee, J.H. and Hong, J.K. (2022) Comparative Performance Analysis of Vibration Prediction Using RNN Techniques. Electronics, 11, Article No. 3619.
https://doi.org/10.3390/electronics11213619
[34] Samaan, G.H., Wadie, A.R., Attia, A.K., Asaad, A.M., Kamel, A.E., Slim, S.O., Abdallah, M.S. and Cho, Y.-I. (2022) MediaPipe’s Landmarks with RNN for Dynamic Sign Language Recognition. Electronics, 11, Article No. 3228.
https://doi.org/10.3390/electronics11193228
[35] Zhu, X., Han, Y., Li, S. and Wang, X. (2022) A Spa-tial-Temporal Topic Model with Sparse Prior and RNN Prior for Bursty Topic Discovering in Social Networks. Journal of Intelligent & Fuzzy Systems, 42, 3909-3922.
https://doi.org/10.3233/JIFS-212135
[36] Hsu, Y.-L., Yang, Y.-T., Wang, J.-S. and Hsu, C.-Y. (2013) Automatic Sleep Stage Recurrent Neural Classifier Using Energy Features of EEG Signals. Neurocomputing, 104, 105-114.
https://doi.org/10.1016/j.neucom.2012.11.003
[37] You, Y., Zhong, X., Liu, G. and Yang, Z. (2022) Automatic Sleep Stage Classification: A Light and Efficient Deep Neural Network Model Based on Time, Frequency and Fractional Fourier Transform Domain Features. Artificial Intelligence in Medicine, 127, Article ID: 102279.
https://doi.org/10.1016/j.artmed.2022.102279
[38] Supratak, A., Dong, H., Wu, C. and Guo, Y. (2017) DeepSleepNet: A Model for Automatic Sleep Stage Scoring Based on Raw Single-Channel EEG. IEEE Transactions on Neural Systems and Rehabilitation Engineering, 25, 1998-2008.
https://doi.org/10.1109/TNSRE.2017.2721116
[39] Fu, M., Wang, Y., Chen, Z., et al. (2021) Deep Learning in Automatic Sleep Staging With a Single Channel Electroencephalography. Frontiers in Physiology, 12, Article 628502.
https://doi.org/10.3389/fphys.2021.628502
[40] Zhuang, L., Dai, M., Zhou, Y. and Sun, L. (2011) Intelligent Au-tomatic Sleep Staging Model Based on CNN and LSTM. Frontiers in Public Health, 10, Article 946833.
https://doi.org/10.3389/fpubh.2022.946833
[41] Zhou, W., Zhu, H., Shen, N., et al. (2011) A Lightweight Seg-mented Attention Network for Sleep Staging by Fusing Local Characteristics and Adjacent Information. IEEE Transac-tions on Neural Systems and Rehabilitation Engineering.
https://doi.org/10.1109/TNSRE.2022.3220372
[42] Chang, R.B. (2022) A Journey toward Artificial Intelli-gence-Assisted Automated Sleep Scoring. Patterns, 3, Article ID: 100429.
https://doi.org/10.1016/j.patter.2021.100429
[43] Casciola, A.A., Carlucci, S.K., Kent, B.A., et al. (2021) A Deep Learning Strategy for Automatic Sleep Staging Based on Two-Channel EEG Headband Data. Sensors, 21, Article No. 3316.
https://doi.org/10.3390/s21103316
[44] Zhao, D., Jiang, R., Feng, M., et al. (2022) A Deep Learning Algo-rithm Based on 1D CNN-LSTM for Automatic Sleep Staging. Technology and Health Care, Preprint, 1-14.
https://doi.org/10.3233/THC-212847
[45] Li, T., Zhang, B., Lv, H., et al. CAttSleepNet: Automatic End-to-End Sleep Staging Using Attention-Based Deep Neural Networks on Single-Channel EEG. International Journal of Envi-ronmental Research and Public Health, 19, Article No. 5199.
https://doi.org/10.3390/ijerph19095199