%0 Journal Article
%T 睡眠自动分期方法综述<br> A Review of Automatic Sleep Staging
%A 徐富献
%A 王抒伟
%A 钱镶钰
%A 胡桓
%A 何情祖
%A 林海
%A 帅建伟
%J Biophysics
%P 34-48
%@ 2330-1694
%D 2019
%I Hans Publishing
%R 10.12677/BIPHY.2019.73004
%X <div style=\"text-align:justify;\">
	<span style=\"font-size:14px;\">睡眠是人体必不可少的一项生理活动。通常，专家将病人整夜的脑电EEG数据以30秒为一帧进行睡眠状态分期并据此进行睡眠状态的分析与评估。然而，依靠人工标记睡眠数据需要消耗大量的精力。另一方面，专家的主观判断也会对分期结果带来误差。所以睡眠的自动分期就变得很重要，本文将介绍近年来的睡眠分期方法，分别是基于统计规则分期方法与基于深度学习技术的分期方法。在统计的分期方法中，介绍了三个重要的过程，预处理、特征提取以及分类器的选择。在基于深度学习的分期方法中，介绍了多层神经网络、卷积神经网络、长短时记忆神经网络以及不同网络组合的神经网络。最后我们对睡眠分期的研究进行了讨论, 认为深度神经网络将是未来睡眠分期研究主要方法。<br />
Sleep is an essential physiological
activity of the human body. Traditionally, doctors divide an entire night's EEG
data into 30-second frames to assess and analyze sleep patterns. However,
manual methods are a huge drain on doctors’ time and energy. And it cannot
avoid the errors of the doctor’s subjectivity. Therefore, the automatic sleep
staging becomes very important. This paper will introduce the sleep staging
methods in recent years, which are based on statistical rules and deep learning
technology respectively. In the statistical staging method, three important
processes including preprocessing, feature extraction and classifier selection,
are introduced. In the staging method based on deep learning, multi-layer
neural networks, convolutional neural networks, long-short term memory neural
networks and neural networks with different combinations are introduced. Finally,
we discuss the sleep staging study and believe that deep neural network will be
the main method to study sleep staging in the future.</span> 
</div>
%K 睡眠自动分期，睡眠脑电，深度学习<br> Automatic Sleep Stage
%K EEG
%K Deep Learning
%U http://www.hanspub.org/journal/PaperInformation.aspx?PaperID=31817