这是用户在 2024-8-25 15:21 为 https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8387348/ 保存的双语快照页面，由 沉浸式翻译 提供双语支持。了解如何保存？
Skip to main content 跳至主要内容
U.S. flag

An official website of the United States government

Access keys 访问键 NCBI Homepage NCBI 主页 MyNCBI Homepage MyNCBI 主页 Main Content 主要内容 Main Navigation 主导航
2021; 11: 17178.
2021； 11：17178。
Published online 2021 Aug 25. doi: 10.1038/s41598-021-95749-2 IF: 3.8 Q1
2021 年 8 月 25 日在线发布。doi：10.1038/s41598-021-95749-2IF：3.8 Q1
PMCID: PMC8387348 IF: 3.8 Q1
PMCID：PMC8387348 IF: 3.8 Q1 IF：3.8 Q1
PMID: 34433839 IF: 3.8 Q1  PMID：34433839IF：3.8 Q1

EEG microstate in obstructive sleep apnea patients
阻塞性睡眠呼吸暂停患者脑电图微状态

Xin Xiong,1 Yuyan Ren,1 Shenghan Gao,1 Jianhua Luo,1 Jiangli Liao,1 Chunwu Wang,2 Sanli Yi,1 Ruixiang Liu,3 Yan Xiang,1 and Jianfeng Hecorresponding author1
熊鑫, 1任宇彦, 1高胜汉, 1罗建华 , 1廖江丽, 1 王春武 , 2易三立, 1刘瑞祥, 3艳翔, 1何建峰 corresponding author 1
Author information Article notes Copyright and License information PMC Disclaimer
作者信息文章注释版权和许可信息PMC 免责声明

Abstract 抽象的

Obstructive sleep apnea (OSA) is a common sleep respiratory disease. Previous studies have found that the wakefulness electroencephalogram (EEG) of OSA patients has changed, such as increased EEG power. However, whether the microstates reflecting the transient state of the brain is abnormal is unclear during obstructive hypopnea (OH). We investigated the microstates of sleep EEG in 100 OSA patients. Then correlation analysis was carried out between microstate parameters and EEG markers of sleep disturbance, such as power spectrum, sample entropy and detrended fluctuation analysis (DFA). OSA_OH patients showed that the microstate C increased presence and the microstate D decreased presence compared to OSA_withoutOH patients and controls. The fifth microstate E appeared during N1-OH, but the probability of other microstates transferring to microstate E was small. According to the correlation analysis, OSA_OH patients in N1-OH showed that the microstate D was positively correlated with delta power, and negatively correlated with beta and alpha power; the transition probability of the microstate B → C and E → C was positively correlated with alpha power. In other sleep stages, the microstate parameters were not correlated with power, sample entropy and FDA. We might interpret that the abnormal transition of brain active areas of OSA patients in N1-OH stage leads to abnormal microstates, which might be related to the change of alpha activity in the cortex.
阻塞性睡眠呼吸暂停（OSA）是一种常见的睡眠呼吸系统疾病。此前的研究发现，OSA患者的清醒状态脑电图（EEG）发生了变化，例如脑电图功率增加。然而，在阻塞性低通气（OH）期间，反映大脑短暂状态的微观状态是否异常尚不清楚。我们研究了 100 名 OSA 患者的睡眠脑电图微观状态。然后对微观状态参数与睡眠障碍的脑电标志物，如功率谱、样本熵和去趋势波动分析（DFA）进行相关分析。 OSA_OH 患者显示，与 OSA_withoutOH 患者和对照相比，微状态 C 增加存在，微状态 D 减少存在。第五个微状态E出现在N1-OH期间，但其他微状态转移到微状态E的概率很小。根据相关性分析，N1-OH中的OSA_OH患者表现出微状态D与δ功率正相关，与β、α功率负相关；微观状态B→C和E→C的转变概率与α功率正相关。在其他睡眠阶段，微观状态参数与功率、样本熵和 FDA 不相关。我们可以解释，OSA患者N1-OH期大脑活动区的异常转变导致了微观状态的异常，这可能与皮层α活性的变化有关。

Subject terms: Neuroscience, Diseases of the nervous system, Encephalopathy, Hypoxic-ischaemic encephalopathy
主题术语：神经科学、神经系统疾病、脑病、缺氧缺血性脑病

Introduction 介绍

Obstructive sleep apnea (OSA) is a common sleep disorder. With the increase of obesity rate, its prevalence is also increasing. It is a chronic multisystem disease, which may lead to a variety of acute clinical problems at the same time, including hypertension, cardio cerebral stroke, mechanical infarction, etc. OSA causes repeated airflow interruption and/or reduction due to stenosis of the upper airway during sleep, resulting in fragmentation of sleep and hypoxemia. These affect the neurobehavioral function of OSA patients, for example, the risk of car accident in OSA patients increases by 2–10 times.
阻塞性睡眠呼吸暂停（OSA）是一种常见的睡眠障碍。随着肥胖率的增加，其患病率也在增加 。它是一种慢性多系统疾病，可能同时导致多种急性临床问题，包括高血压、心脑卒中、机械性梗塞等。 OSA 由于上呼吸道狭窄导致气流反复中断和/或减少。睡眠期间呼吸道通畅，导致睡眠碎片化和低氧血症 。这些都会影响OSA患者的神经行为功能，例如OSA患者发生车祸的风险增加2-10倍 。

Despite the prevalence of OSA, its underlying neurophysiological process is unclear. Clinical studies have shown that the EEG of OSA patients has changed, such as changes in power spectrum and energy,. Zhou et al. found that the sample entropy of sleep apnea syndrome patients was lower than healthy controls in each sleep stage. Grenèche et al. found that the power of wakefulness EEG in OSA patients after 24 h of sleep deprivation was higher than healthy controls. D’Rozario et al. found that the wakefulness EEG power spectrum and detrended fluctuation analysis (DFA) of OSA patients were related to simulated driving performance, and the scale index α of DFA can be used as an indicator of simulated driving performance,. Kim et al. used DFA to analyze the sleep onset period (SOP) of narcolepsy patients, and found that the SOP of narcolepsy patients was significantly larger compared with healthy controls.
尽管 OSA 很常见，但其潜在的神经生理过程尚不清楚。临床研究表明，OSA患者的脑电图发生了变化，如功率谱和能量的变化 , 。周等人。研究发现，睡眠呼吸暂停综合征患者的样本熵在每个睡眠阶段均低于健康对照者 .格雷诺什等人。发现 OSA 患者在睡眠剥夺 24 小时后的觉醒脑电图功率高于健康对照 。德罗萨里奥等人。研究发现，OSA患者的清醒状态脑电功率谱和去趋势波动分析（DFA）与模拟驾驶表现相关，DFA的量表指数α可以作为模拟驾驶表现的指标 , 。金等人。利用DFA分析发作性睡病患者的睡眠开始期（SOP），发现发作性睡病患者的SOP明显大于健康对照者 。

These studies provide a window for us to understand OSA patients’ EEG. However, such studies are relatively few and focus on the difference of neurobehavioral ability (such as simulated driving) of OSA patients and healthy subjects–. To the best of our knowledge, there have been no reports of studying the EEG of OSA patients from the perspective of microstate. EEG microstate employs the information of the entire time and space of EEG to characterize the rapid spontaneous change of scalp potential with time. Such this approach can provide a more informative framework and global interpretability without any type of a priori hypothesis, in contrast with other EEG analysis techniques, which evaluate the brain’s electrical field in a specific location (for example by a priori choice of electrodes of interest) or at determinate time intervals or in specific frequency bands. According to the microstate theory, EEG signals are composed of a series of topographic maps with two remarkable properties,: (1) EEG signals can be expressed with a small amount of topographic maps; (2) before the transition from one topographic map to another, a single topographic map dominates with duration of about 80–120 ms. These metastable states are microstates, which are described as the basic components or "thought atoms" of human information processing. Therefore, the microstate analysis method is more used to study human's cognition and thinking, as well as psychotic disorders. However, few researchers use it to study sleep, only the healthy subjects’ and narcolepsy patients’ sleep,. Brodbeck et al. found that healthy subjects had 4 microstates during the wakefulness and NREM sleep stages. Kuhn et al. found that narcoleptic patients in early NREM sleep had an additional microstate E during the N3 phase. Dose the microstates of EEG in OSA patients change? If the microstates change, is there any correlation between microstate parameters and EEG markers such as power spectrum,,,, sample entropy,, and DFA–?
这些研究为我们了解OSA患者的脑电图提供了一个窗口。然而，此类研究相对较少，并且集中于 OSA 患者和健康受试者的神经行为能力（例如模拟驾驶）的差异 – 。据我们所知，目前还没有从微观状态角度研究OSA患者脑电图的报道。脑电微状态利用脑电图整个时间和空间的信息来表征头皮电位随时间的快速自发变化 。与其他脑电图分析技术相比，这种方法可以提供信息更丰富的框架和全局可解释性，而无需任何类型的先验假设 ，其他脑电图分析技术评估特定位置的大脑电场（例如通过先验选择感兴趣的电极） ）或以确定的时间间隔或在特定频带 。根据微观状态理论，脑电信号由一系列地形图组成，具有两个显着的特性 , ： (1)脑电信号可以用少量的地形图来表达； (2) 在从一张地形图转换到另一张地形图之前，单一地形图占主导地位，持续时间约为 80-120 毫秒。这些亚稳态是微观状态，被描述为人类信息处理的基本组成部分或“思想原子”。 因此，微观状态分析方法更多地用于研究人类的认知和思维，以及精神障碍 。然而，很少有研究人员用它来研究睡眠，只有健康受试者和发作性睡病患者的睡眠 , 。布罗贝克等人。发现健康受试者在清醒和 NREM 睡眠阶段有 4 种微观状态 。库恩等人。发现早期 NREM 睡眠中的发作性睡病患者在 N3 阶段有一个额外的微状态 E 。 OSA 患者脑电图的微观状态会发生变化吗？如果微状态发生变化，微状态参数和 EEG 标记（例如 样本和 DFA – ）之间是否存在相关性？

Therefore, we hypothesized that there were abnormal microstates in OSA patients during sleep obstructive apnea or obstructive hypopnea, and the microstate parameters were correlated with power spectrum, sample entropy and DFA.
因此，我们推测OSA患者在睡眠阻塞性呼吸暂停或阻塞性低通气期间存在异常微状态，且微状态参数与功率谱、样本熵和DFA相关。

Results 结果

OSA of sleep stages 睡眠阶段的 OSA

The stages of sleep include waking, non-rapid eye movement (N1, N2, N3), and rapid eye movement (R) . OSA may occur at any time in the sleep cycle, and the number of obstructive hypopnea (OH) occurrences in 4 sleep stages in 100 OSA patients in Subgroup_I was counted. Obstructive hypopnea occurs more frequently in N1, N2 and R stages, less in N3 stage, and least in W stages. Patients with more OH in N1 and N2 stages also have more OH in R stages, such as Sub61 and Sub80. However, patients with more OH in R stages may not have more OH in N1 and N2 stages, such as Sub45, Sub73, Sub87, Sub96, and Sub97.
睡眠阶段包括清醒、非快速眼动（N1、N2、N3）和快速眼动（R） 。 OSA可能发生在睡眠周期的任何时间，统计了Subgroup_I中100名OSA患者在4个睡眠阶段发生阻塞性低通气（OH）的次数。阻塞性呼吸不足在N1、N2和R期发生频率较高，N3期较少，W期最少。 N1 和 N2 期 OH 较多的患者在 R 期（例如 Sub61 和 Sub80）也有较多 OH。然而，R 期 OH 较多的患者在 N1 和 N2 期可能不会出现较多 OH，例如 Sub45、Sub73、Sub87、Sub96 和 Sub97。

OSA microstates OSA 微观状态

We used CARTOOL software to estimate the microstates in OSA_OH (OSA with obstructive hypopnea) patients and controls, as shown in Fig. 1. The controls have four similar microstates A, B, C and D in N1, N2, N3 and R stages (ignoring the polarity of microstates,, which is similar to the previously reported microstate of sleep EEG,. The fifth microstate E of OSA_OH appears in N1_OH and N3_OH stages. Since there was no corresponding microstate E in controls, we made a comparison in two parts :(1) 4 typical microstates A, B, C and D were estimated respectively in OSA_OH patients and controls; (2) 5 microstates A, B, C, D and E found in N1-OH were used as topographic map templates, which were employed to fit all the sleep stages in OSA patients with and without obstructive hypopnea, that is, 5 microstates A,B,C and D and E were estimated in OSA_OH (OSA with obstructive hypopnea) and OSA_withoutOH (OSA without obstructive hypopnea) patients. Four microstates explained 71.7%, 73.4%, 76.4%, and 72.3% of the global variances in four sleep stages in OSA_OH patients, while five microstates explained 76.1%, 74.2%, 76.5%, and 72.1% of the global variances.
我们使用CARTOOL软件来估计OSA_OH（伴有阻塞性低通气的OSA）患者和对照的微观状态，如图1所示。对照在 N1、N2、N3 和 R 阶段有四个相似的微状态 A、B、C 和 D（忽略微的极性，这与之前报道的睡眠脑电图微类似。第五个微状态OSA_OH的E出现在N1_OH和N3_OH阶段，由于对照中没有相应的微状态E，我们分两部分进行比较：（1）分别在OSA_OH患者和对照中估计4种典型的微状态A、B、C和D； (2)以N1-OH中发现的5个微状态A、B、C、D和E作为地形图模板，用于拟合有和没有阻塞性低通气的OSA患者的所有睡眠阶段，即5个微状态A在 OSA_OH（伴有阻塞性呼吸不足的 OSA）和 OSA_withoutOH（不伴有阻塞性呼吸不足的 OSA）患者中估计了 、B、C、D 和 E，四种微观状态解释了四种睡眠中 71.7%、73.4%、76.4% 和 72.3% 的总体差异。 OSA_OH 患者的各个阶段，而五种微观状态解释了全局差异的 76.1%、74.2%、76.5% 和 72.1%。

An external file that holds a picture, illustration, etc. Object name is 41598_2021_95749_Fig1_HTML.jpg

Microstates in W, N1, N2, N3 and R sleep stages in controls and OSA_OH patients.
对照组和 OSA_OH 患者 W、N1、N2、N3 和 R 睡眠阶段的微观状态。

The calculated microstate parameters include: Global Explained Variance (Gev (%)), Mean Duration (MD (ms)), Time Coverage (TC (%)), Occurrence (Oc (/s)), and Transition Probability (TP).
计算的微观状态参数包括：全局解释方差 (Gev (%))、平均持续时间 (MD (ms))、时间覆盖 (TC (%))、发生率 (Oc (/s)) 和转移概率 (TP)。

The 2 × 4 × 4 (2 groups, 4 microstates, 4 microstate parameters (Gev, MD, TC, Oc)) and 2 × 5 × 4 multivariate ANOVA for each sleep stage was performed for OSA_OH vs. Control and OSA_OH vs. OSA_withoutOH, respectively, and the Bonferroni post-hoc tests were performed. The total correction models of OSA_OH vs. Control (N1: F(31,928) = 19.283, P = 0.000; N2: F(31,944) = 28.405, P = 0.000; N3: F(31,576) = 11.915, P = 0.000; R: F(31,880) = 20.430, P = 0.000) and OSA_OH vs. OSA_withoutOH (N1: F(39,2280) = 38.047, P = 0.000; N2: F(39,2296) = 28.104, P = 0.000; N3: F(39,1938) = 27.316, P = 0.000; R: F(39,2232) = 28.407, P = 0.000) showed significant differences. Then, a two-sided one-sample t test was performed on the microstate parameters within group, with significant differences indicated by stars in Tables ​Tables1,1, ​,22 and ​and33.
对 OSA_OH 与对照组以及 OSA_OH 与 OSA_withoutOH 进行每个睡眠阶段的 2 × 4 × 4（2 组、4 个微状态、4 个微状态参数（Gev、MD、TC、Oc））和 2 × 5 × 4 多元方差分析分别进行了 Bonferroni 事后测试。 OSA_OH vs. Control 的总校正模型 (N1: F(31,928) = 19.283, P = 0.000; N2: F(31,944) = 28.405, P = 0.000; N3: F(31,576) = 11.915, P = 0.000; R ：F(31,880) = 20.430，P = 0.000) 和 OSA_OH 与 OSA_withoutOH 对比 (N1：F(39,2280) = 38.047，P = 0.000；N2：F(39,2296) = 28.104，P = 0.000；N3： F(39,1938) = 27.316, P = 0.000; R: F(39,2232) = 28.407, P = 0.000) 显示显着差异。然后，对组内微观状态参数进行双边单样本t检验，表​表1、1 、 ​、2 2和​和3 3中用星号表示显着差异。

Table 1 表格1

Microstate parameters of 4 microstates in controls in 4 sleep stages.
4 个睡眠阶段对照中 4 个微状态的微状态参数。

Gev (%) 热值 (%)MD (ms) MD（毫秒）TC (%)Oc/s 奥克/秒
Average 平均的SDAverage 平均的SDAverage 平均的SDAverage 平均的SD
N1
A16.30.735.021.124.10.44.3*0.3
B16.8*1.035.624.325.50.64.40.4
C26.4**0.535.4**27.124.4**0.34.3*0.3
D17.00.934.324.024.1**0.34.3**0.3
N2
A17.00.639.026.124.30.43.90.3
B18.20.640.325.026.1**0.54.1*0.3
C28.2**0.740.6*26.726.0**0.34.00.3
D16.11.035.922.820.90.53.70.3
N3
A18.51.254.5108.024.7**0.83.00.4
B19.21.055.3107.325.40.43.10.5
C29.3*0.555.898.925.3*0.43.10.4
D16.11.049.1*98.320.40.62.80.4
R
A17.71.238.831.825.3*0.54.0*0.3
B17.31.238.629.125.20.84.10.4
C27.9*0.439.732.026.2*0.54.10.4
D15.71.435.228.820.8*0.53.7*0.2

The stars indicate the significant difference at each sleep stage between controls and OSA-OH patients (*P < 0.05, **P < 0.01).
星星表示对照组和 OSA-OH 患者在每个睡眠阶段的显着差异（*P < 0.05，**P < 0.01）。

Table 2 表2

Microstate parameters of 4 microstates in OSA-OH patients in 4 sleep stages.
OSA-OH患者4个睡眠阶段4个微状态的微状态参数。

Gev (%) 热值 (%)MD (ms) MD（毫秒）TC (%)Oc/s 奥克/秒
Average 平均的SDAverage 平均的SDAverage 平均的SDAverage 平均的SD
N1-OH
A15.95.435.33.922.84.54.00.7
B19.89.637.453.125.95.94.2*
4.2 *		0.5
C20.9*
20.9 *		7.339.0*
39.0 *		50.027.7**
27.7 **		4.74.30.6
D15.15.234.0*42.021.14.03.8*0.6
N2-OH
A16.64.139.043.323.53.43.80.5
B17.24.240.047.024.23.13.80.5
C22.7**9.243.158.728.5**5.74.0*0.5
D16.910.437.254.020.9*4.53.5*0.6
N3-OH
A16.26.150.798.820.9*5.82.60.7
B21.27.959.0221.126.77.03.00.5
C22.46.858.5106.727.3*5.13.00.5
D16.67.450.4131.121.15.02.70.6
R-OH
A16.34.838.440.523.84.23.80.6
B19.610.832.345.124.95.13.9*0.5
C21.9*10.341.9*52.328.2*5.84.10.5
D14.53.735.534.120.02.63.5*0.4

The stars indicate the significant difference at each sleep stage between OSA-OH and OSA-withoutOH patients (*P < 0.05, **P < 0.01).
星号表示 OSA-OH 和 OSA-withoutOH 患者在每个睡眠阶段的显着差异（*P < 0.05，**P < 0.01）。

Table 3 表3

Microstate parameters of 5 microstates in OSA-OH patients in 4 sleep stages.
OSA-OH患者4个睡眠阶段5种微状态的微状态参数。

Gev (%) 热值 (%)MD (ms) MD（毫秒）TC (%)Oc/s 奥克/秒
Average 平均的SDAverage 平均的SDAverage 平均的SDAverage 平均的SD
N1-OH
A12.34.632.733.517.23.43.3*0.6
B20.310.037.9*54.226.86.14.20.5
C19.5**7.337.9**40.926.1**4.64.20.5
D15.05.133.938.821.2*4.03.9*0.6
E9.0**2.428.435.26.7**2.01.6**0.2
N2-OH
A12.53.735.436.117.52.73.20.5
B17.24.140.142.724.33.13.80.5
C20.7**9.341.256.026.1**5.73.9*0.5
D17.310.337.553.222.04.53.60.6
E6.53.730.740.27.8*2.81.60.4
N3-OH
A15.94.949.797.420.04.12.60.4
B19.46.053.995.724.85.22.90.5
C21.9*6.356.689.027.0*4.83.00.5
D16.37.148.191.221.05.12.70.5
E3.31.844.5117.06.5**2.81.0**0.4
R-OH
A13.8*3.736.439.420.13.43.40.5
B19.410.539.445.624.85.13.80.5
C20.810.641.449.527.2*5.94.00.5
D14.6*3.736.035.420.7*2.93.60.5
E3.52.033.068.06.8*2.21.3*0.4

The stars indicate the significant difference at each sleep stage between OSA-OH and OSA-withoutOH patients (*P < 0.05, **P < 0.01). The bold value has significant difference, *P < 0.05, **P < 0.005.
星号表示 OSA-OH 和 OSA-withoutOH 患者在每个睡眠阶段的显着差异（*P < 0.05，**P < 0.01）。粗体值有显着性差异，*P < 0.05，**P < 0.005。

In Table ​Table1,1, there were significant differences in parameters of microstate A, B, C and D in the four sleep stages between OSA_OH patients and controls. In Table ​Table2,2, BOC, CGev, CMD, CTC, DMD, DOC in N1-OH, CGev, CTC, COC, DTC, DOC in N2-OH, ATC, CTC in N3-OH, BOC, CGev, CMD,CTC, DMD, DOC in R-OH were significantly different from those of OSA_withouOH patients. The test results of transition probability were shown in Appendix. This indicated that the microstates C and D in OSA_OH patients changed with the typical four microstate states as the fitting template, and the CTC of OSA_OH was larger than that of OSA_withoueOH (N1: 0.28 ± 0.05 vs. 0.25 ± 0.01; N2: 0.29 ± 0.06 vs. 0.25 ± 0.00; N3: 0.27 ± 0.05 vs. 0.26 ± 0.00; R: 0.28 ± 0.05 vs. 0.26 ± 0.01), and the DOC of OSA_OH was smaller than that of OSA_withoueOH (N1: 3.8 ± 0.01 vs. 4.0 ± 0.00; N2: 3.5 ± 0.01 vs. 3.7 ± 0.00; N3: 3.5 ± 0.01 vs. 3.7 ± 0.00; R: 3.8 ± 0.01 vs. 4.0 ± 0.00). At the same time, transition probability TPA→C, TPB→C, TPA→D, TPB→D and TPD→C had significant differences in the four sleep stages. As shown in Fig. 2, the transition probability of microstates A → C, B → C, A → D, B → D, D → C in all sleep stages were relatively large.
在表​表1、1中，OSA_OH患者和对照组在四个睡眠阶段的微状态A、B、C和D参数存在显着差异。表2中，B OC, C Gev , C MD , C TC , D MD , D OC in N1-OH, C Gev , C TC , C OC , D TC , D OC in N2-OH, A TC N3-OH 中的 、C TC ，R-OH 中的 B OC 、 C Gev 、C MD 、C TC 、D MD 、D OC与 OSA_withouOH 患者有显着差异。转移概率的检验结果见附录。这表明 OSA_OH 患者的微状态 C 和 D 以典型的四种微状态为拟合模板发生变化，并且 OSA_OH 的 C TC 大于 OSA_withoueOH 的 C TC (N1: 0.28 ± 0.05 vs 0.25 ± 0.01; N2: 0.29 ± 0.06与0.25 ± 0.00；N3：0.27 ± 0.05与0.26 ± 0.00；R：0.28 ± 0.05与0.26 ± 0.01），且 OSA_OH 的 D OC小于 OSA_withoueOH（N1：3.8 ± 0.01） 4.0±0.00； N2 ：3.5±0.01 ； N3 ：3.5±0.01； R ：3.8±0.01。 同时，四个睡眠阶段的转移概率TP A→C 、TP B→C 、TP A→D 、TP B→D和TP D→C存在显着差异。如图2所示，各个睡眠阶段微观状态A→C、B→C、A→D、B→D、D→C的转变概率都较大。

An external file that holds a picture, illustration, etc. Object name is 41598_2021_95749_Fig2_HTML.jpg

Transition probabilities (TP) of 4 microstates in controls, OSA_OH and OSA_withoutOH patients.
对照组、OSA_OH 和 OSA_withoutOH 患者中 4 种微状态的转移概率 (TP)。

In Table ​Table3,3, although microstate E was only found in OSA _OH patients in N1-OH and N3- OH stages through conventional microstate calculation process, we found that there was a small amount of microstate E in each sleep stage after fitting all the sleep stages with the five microstates. Microstate E accounted for larger global variance (9%) in N1-OH. However, compared with microstates A, B, C and D during all sleep stages, microstates E occurred for shorter period of time and accounted for lower proportion of EEG signals.
在表​表3、3中，虽然通过传统的微状态计算过程仅在OSA_OH患者的N1-OH和N3-OH阶段发现了微状态E，但在拟合所有数据后，我们发现每个睡眠阶段都存在少量的微状态E。睡眠阶段的五个微观状态。 Microstate E 解释了 N1-OH 中较大的全局方差 (9%)。然而，与所有睡眠阶段的微状态A、B、C和D相比，微状态E出现的时间较短，并且占脑电信号的比例较低。

In Table ​Table3,3, the AOC, BMD, CGev, CMD, CTC, DTC, DOC, EGev, ETC, EOC in N1-OH, CGev, CTC, COC, ETC in N2-OH, CGev, CTC, ETC, EOC in N3-OH, AGev, CTC, DGev, DTC, EOC in R-OH were significantly different from those of OSA_withoutOH patients. Among these parameters, it was mainly CTC and ETC that had changed. Compared with the results of the four microstates in Table ​Table2,2, CTC of the four sleep stages were all smaller, which indicated that part of the EEG segments that were originally labeled as microstate C were labeled as microstate E or other unlabeled microstates. Therefore, CTC and ETC had significant differences. However, the CTC of OSA_OH was still larger than that of OSA_withoutOH, while the ETC and ESD of OSA_OH were smaller than that of OSA_withoutOH, as shown in Table ​Table4.4. The transition probability from microstates A, B, D to C, E, and from microstate E to C all had significant differences, in which the transition probability from microstates A, B, D, E to C was much greater than that of E (0.31 vs. 0.10), as shown in Table ​Table55 (only part of the transition probabilities with significant differences were given). And the transition probability from microstates A, B, D to C was greater than that of OSA_withoutOH, while the transition probability from microstates A, B, C, D to E was less than that of OSA_withoutOH, as shown in Fig. 3. This indicated that the microstates A, B, C, D and E transferred to microstates C with a high probability, and to E with a low probability after the occurrence of obstructive hypopnea.
表3、3中，N1-OH中的A OC , B MD , C Gev , C MD , C TC , D TC , D OC , E Gev , E TC , E OC , C Gev , C TC , C OC N2-OH 中的 E TC ，N3-OH 中的 C Gev 、 C TC 、 E TC 、 E OC ，R-OH 中的 A Gev 、 C TC 、 D Gev 、 D TC 、 E OC与 OSA_withoutOH 患者有显着差异。这些参数中，主要是C TC和E TC发生了变化。与表​表2、2中四个微状态的结果相比，四个睡眠阶段的C TC均较小，这表明部分原本标记为微状态C的脑电片段被标记为微状态E或其他未标记的微状态。因此，C TC和E TC具有显着差异。然而，OSA_OH的C TC仍然大于OSA_withoutOH，而OSA_OH的E TC和E SD小于OSA_withoutOH，如表4所示。 4 . 微状态A、B、D到C、E以及微状态E到C的转移概率均存在显着差异，其中微状态A、B、D、E到C的转移概率远大于E（ 0.31 vs. 0.10），如表5所示（仅给出了差异显着的部分转移概率）。并且从微状态A、B、D到C的转移概率大于OSA_withoutOH，而从微状态A、B、C、D到E的转移概率小于OSA_withoutOH，如图3所示。这表明，阻塞性低通气发生后，微状态A、B、C、D、E大概率转移至微状态C，小概率转移至E。

Table 4 表4

Significantly different microstate parameters CTC, ETC, Eoc of 5 microstates in OSA_OH (marked 1 in the table) and OSA_withoutOH (marked 0 in the table) patients.
OSA_OH（表中标记1）和OSA_withoutOH（表中标记0）患者5个微状态的微状态参数C TC 、E TC 、Eoc显着不同。

Sleep stage 睡眠阶段Groups 团体CTCETCEoc 奥克
N110.26 ± 0.05 0.26±0.050.07 ± 0.02 0.07±0.020.02 ± 0.00 0.02±0.00
00.24 ± 0.01 0.24±0.010.08 ± 0.02 0.08±0.020.02 ± 0.00 0.02±0.00
N210.26 ± 0.06 0.26±0.060.08 ± 0.03 0.08±0.030.02 ± 0.00 0.02±0.00
00.23 ± 0.01 0.23±0.010.09 ± 0.01 0.09±0.010.02 ± 0.00 0.02±0.00
N310.27 ± 0.05 0.27±0.050.06 ± 0.03 0.06±0.030.01 ± 0.00 0.01±0.00
00.24 ± 0.00 0.24±0.000.08 ± 0.01 0.08±0.010.01 ± 0.00 0.01±0.00
R10.27 ± 0.06 0.27±0.060.07 ± 0.02 0.07±0.020.01 ± 0.00 0.01±0.00
00.25 ± 0.01 0.25±0.010.07 ± 0.01 0.07±0.010.01 ± 0.00 0.01±0.00

Table 5 表5

Significantly different transition probabilities of 5 microstates in OSA_OH and OSA_withoutOH patients (– indicates no significance in the table).
OSA_OH 和 OSA_withoutOH 患者中 5 种微状态的转变概率显着不同（– 表示表中无显着性）。

TPA→C TP A→CTPA→E TP A→ETPB→C TP B→CTPB→E TP B→ETPC→E TP C→ETPD→C TP D→CTPD→E TP D→ETPE→C TP E→C
N1
10.31 ± 0.05 0.31±0.050.11 ± 0.03 0.11±0.030.39 ± 0.07 0.39±0.070.08 ± 0.03 0.08±0.030.12 ± 0.03 0.12±0.030.24 ± 0.05 0.24±0.050.08 ± 0.02 0.08±0.020.32 ± 0.06 0.32±0.06
00.28 ± 0.01 0.28±0.010.13 ± 0.02 0.13±0.020.34 ± 0.00 0.34±0.000.09 ± 0.01 0.09±0.010.13 ± 0.02 0.13±0.020.22 ± 0.02 0.22±0.020.10 ± 0.01 0.10±0.010.30 ± 0.01 0.30±0.01
N2
10.31 ± 0.06 0.31±0.060.12 ± 0.04 0.12±0.040.37 ± 0.07 0.37±0.070.09 ± 0.03 0.09±0.030.13 ± 0.04 0.13±0.040.24 ± 0.06 0.24±0.060.09 ± 0.03 0.09±0.030.32 ± 0.05 0.32±0.05
00.28 ± 0.00 0.28±0.000.14 ± 0.02 0.14±0.020.33 ± 0.00 0.33±0.000.10 ± 0.00 0.10±0.000.14 ± 0.02 0.14±0.020.22 ± 0.02 0.22±0.020.10 ± 0.01 0.10±0.010.29 ± 0.01 0.29±0.01
N3
10.38 ± 0.07 0.38±0.07–0.40 ± 0.09 0.40±0.09–0.09 ± 0.03 0.09±0.030.22 ± 0.05 0.22±0.050.08 ± 0.03 0.08±0.030.32 ± 0.17 0.32±0.17
00.35 ± 0.01 0.35±0.01–0.36 ± 0.01 0.36±0.01–0.10 ± 0.01 0.10±0.010.19 ± 0.02 0.19±0.020.10 ± 0.01 0.10±0.010.24 ± 0.01 0.24±0.01
R
1––0.38 ± 0.06 0.38±0.060.09 ± 0.02 0.09±0.02–0.24 ± 0.04 0.24±0.04––
0––0.36 ± 0.010.10 ± 0.01–0.23 ± 0.02––
An external file that holds a picture, illustration, etc. Object name is 41598_2021_95749_Fig3_HTML.jpg

Transition probabilities (TP) of 5 microstates in OSA_OH and OSA_withoutOH patients.
OSA_OH 和 OSA_withoutOH 患者 5 种微状态的转变概率 (TP)。

Power, sample entropy and DFA of OSA patients
OSA 患者的功效、样本熵和 DFA

The power, sample entropy and DFA of OSA_OH vs. Control and OSA_OH vs. OSA_withoutOH were respectively analyzed by 2 × 7 multivariate ANOVA, showing significant differences. Then, a two-sided one-sample t test was conducted within group, as shown in Figs. 4 and ​and55.
分别采用2×7多元方差分析对OSA_OH vs. Control和OSA_OH vs. OSA_withoutOH的功效、样本熵和DFA进行分析，显示出显着差异。然后进行组内双边单样本t检验，如图1、2所示。 4和​和 5 5 。

An external file that holds a picture, illustration, etc. Object name is 41598_2021_95749_Fig4_HTML.jpg

The delta, theta, alpha, sigma and beta power in four sleep stages in controls, OSA_OH and OSA_withoutOH patients ( P* < 0.05, P** < 0.05).
对照组、OSA_OH 和 OSA_withoutOH 患者四个睡眠阶段的 delta、theta、alpha、sigma 和 beta 功率（P* < 0.05，P** < 0.05）。

An external file that holds a picture, illustration, etc. Object name is 41598_2021_95749_Fig5_HTML.jpg

Sample entropy and FDA in four sleep stages in controls, OSA_OH and OSA_withoutOH patients (P* < 0.05, P** < 0.05).
对照组、OSA_OH 和 OSA_withoutOH 患者四个睡眠阶段的样本熵和 FDA (P* < 0.05，P** < 0.05)。

For OSA_OH vs. control, there was a significant difference in beta power in N1 (P = 0.026), in sigma power in N2 (P = 0.02), in delta, theta and alpha power in N3 stage (P = 0.021; P = 0.027; P = 0.007), in beta power in R (P = 0.021). For OSA_OH vs. OSA_withoutOH, there was a significant difference in beta power in N2 (P = 0.02), in beta power in N3 (P = 0.04), in theta power in R (P = 0.016).
对于 OSA_OH 与对照，N1 阶段的 β 功率（P = 0.026）、N2 阶段的 sigma 功率（P = 0.02）、N3 阶段的 delta、theta 和 alpha 功率（P = 0.021；P = 0.027；P = 0.007），R 的 beta 幂（P = 0.021）。对于 OSA_OH 与 OSA_withoutOH，N2 的 β 功率 (P = 0.02)、N3 的 β 功率 (P = 0.04)、R 的 θ 功率 (P = 0.016) 存在显着差异。

The sample entropy of OSA_OH vs. control showed significant difference only in N1 (P = 0.015), while FDA showed significant difference in N1, N2, N3, R (P = 0.005; P = 0.001; P = 0.005; P = 0.005). The sample entropy of OSA_OH vs. OSA_withoutOH showed significant difference only in N1, N2 (P = 0.043; P = 0.042), and there was no significant difference in FDA.
OSA_OH与对照的样本熵仅在N1中表现出显着差异（P = 0.015），而FDA在N1、N2、N3、R中表现出显着差异（P = 0.005；P = 0.001；P = 0.005；P = 0.005） 。 OSA_OH与OSA_withoutOH的样本熵仅在N1、N2中表现出显着差异（P = 0.043；P = 0.042），而在FDA中没有显着差异。

Correlation between microstate parameters and power, sample entropy and DFA
微观状态参数与功率、样本熵和DFA之间的相关性

The correlation between the parameters of 4 and 5 microstates and power, sample entropy and DFA in OSA_OH and OSA_withoutOH patients was respectively compared, as shown in Tables ​Tables66 and ​and77.
分别比较OSA_OH和OSA_withoutOH患者的4和5微状态参数与功率、样本熵和DFA的相关性，如表​表6和​和7 7所示。

Table 6 表6

Correlation between microstate parameters of 4 microstates and power, sample entropy and DFA in OSA_OH and OSA_withoutOH patients (P < 0.05/7, i.e., 7 being the number of microstates and power, sample entropy and DFA).
OSA_OH和OSA_withoutOH患者4种微状态的微状态参数与功率、样本熵和DFA之间的相关性(P < 0.05/7，即7为微状态数和功率、样本熵和DFA)。

OSA_OH (4 microstates) OSA_OH（4 个微观状态）OSA_withoutOH (4 microstates)
OSA_withoutOH（4 个微观状态）
Correlation 相关性Correlation 相关性
N1DMD − P_beta (r = − 0.362, P = 0.001)
D MD – P_beta (r = – 0.362, P = 0.001)		N1No correlation 没有相关性
DTC − P_beta (r = − 0.365, P = 0.007)
D TC − P_beta (r = − 0.365，P = 0.007)
DOC − P_delta (r = 0.380, P = 0.007)
D OC – P_delta（r = 0.380，P = 0.007）
DOC − P_alpha (r = − 0.382, P = 0.006)
D OC – P_alpha (r = – 0.382, P = 0.006)
TPB→C − P_alpha (r = 0.388, P = 0.005)
TP B→C − P_alpha (r = 0.388, P = 0.005)
N2No correlation 没有相关性N2No correlation 没有相关性
N3No correlation 没有相关性N3No correlation 没有相关性
RNo correlation 没有相关性RNo correlation 没有相关性

Table 7 表7

Correlation between microstate parameters of 5 microstates and power, sample entropy and DFA in OSA_OH and OSA_withoutOH patients.
OSA_OH和OSA_withoutOH患者5种微状态的微状态参数与功率、样本熵和DFA的相关性。

OSA_OH (5 microstates) OSA_OH（5 个微观状态）OSA_withoutOH (5 microstates)
OSA_withoutOH（5 个微观状态）
Correlation 相关性Correlation 相关性
N1DOC − P_delta (r = 0.401, P = 0.004)
D OC – P_delta（r = 0.401，P = 0.004）		N1TPX→Y − P_delta, P_theta
TP X→Y − P_delta, P_theta
DOC − P_alpha (r = − 0.376, P = 0.007)
D OC – P_alpha (r = – 0.376, P = 0.007)		TPX→Y − SE
TP X→Y − SE
TPB→C − P_alpha (r = 0.486, P = 0.000)
TP B→C − P_alpha (r = 0.486, P = 0.000)		TPX→Y − FDA
TP X→Y - FDA
TPE→C − P_alpha (r = 0.412, P = 0.003)
TP E→C − P_alpha (r = 0.412, P = 0.003)
N2TPA→B − FDA(r = − 0.364, P = 0.009)
TP A→B − FDA(r = − 0.364, P = 0.009)		N2TPX→Y − P_delta, P_theta, P_alpha
TP X→Y − P_delta、P_theta、P_alpha
TPA→E − FDA(r = 0.396, P = 0.004)
TP A→E − FDA(r = 0.396, P = 0.004)		TPX→Y − FDA
TP X→Y - FDA
N3TPA→E − FDA(r = − 0.490, P = 0.009)N3TPX→Y − P_sigma
RTPC→E − FDA(r = 0.462, P = 0.001)RTPX→Y − P_delta, P_theta, P_alpha,P_sigma,P_beta
TPX→Y − SE
TPX→Y − FDA

In Table ​Table6,6, OSA_ withoutOH had no correlation in the four sleep stages, while OSA_OH had correlation only in N1, and only the parameters of microstate D was correlated with delta, alpha and beta power. They were positively correlated with delta power (r = 0.380, P = 0.007), and negatively correlated with alpha and beta power (r = − 0.362, P = 0.001; R = − 0.382, P = 0.006), and the TPB→C was positively correlated with alpha power (r = 0.388, P = 0.005).
表​表6、6中，OSA_withoutOH在四个睡眠阶段没有相关性，而OSA_OH仅在N1中具有相关性，并且只有微状态D的参数与delta、alpha和beta功率相关。它们与 delta 功率呈正相关（r = 0.380，P = 0.007），与 alpha 和 beta 功率呈负相关（r = − 0.362，P = 0.001；R = − 0.382，P = 0.006），并且 TP B→ C与 alpha 幂呈正相关（r = 0.388，P = 0.005）。

In Table ​Table7,7, only parameters of microstate D in OSA_OH were correlated with delta and alpha power in N1, and the TPB→C and TPE→C was positively correlated with alpha power. The TPA→B and TPA→E were related to FDA in N2; TPA→E were related to FDA in N3; TPC→E were related to FDA in R. The five microstate parameters in OSA_withoutOH patients were all not correlated in four sleep stages, but various transition probabilities (TPX→Y) were correlated with delta and theta power and sample entropy (SE) and FDA in N1; TPX→Y were correlated with delta, theta, alpha power and FDA in N2; TPX→Y were only correlated with sigma power in N3; TPX→Y were correlated with delta, theta, alpha, sigma, beta power and SE and FDA in R.
在表7、7中，只有OSA_OH中微状态D的参数与N1中的δ和α功率相关，并且TP B→C和TP E→C与α功率正相关。 N2中TP A→B和TP A→E与FDA相关； N3中TP A→E与FDA相关； TP C→E与 R 中的 FDA 相关。OSA_withoutOH 患者的五个微观状态参数在四个睡眠阶段均不相关，但各种转换概率（TP X→Y ）与 delta 和 theta 功率以及样本熵（SE）相关和 FDA 的 N1 级； TP X→Y与 N2 中的 delta、theta、alpha 幂和 FDA 相关； TP X→Y仅与 N3 中的 sigma 幂相关； TP X→Y与 R 中的 delta、theta、alpha、sigma、beta 幂以及 SE 和 FDA 相关。

Discussion 讨论

Microstate reflects the instantaneous state of the brain, and can identify discontinuous and nonlinear changes of global functional brain state under very high temporal resolution,. It has been found that four canonical microstates A, B, C and D are related to the activities of the posterior cingulate cortex. Brodbeck et al. investigated wakefulness and NREM sleep of healthy subjects, and found that microstate C was dominant in W, N1 and N3 stages, while microstate B was dominant in N2 stage. With the increase of sleep depth, the parameter GEV of microstate D gradually decreased. Kuhn et al. investigated early NREM sleep of narcoleptic patients, and found that microstate C and D were dominant in N1 stage, microstate D was still dominant in N2 and N3 stages, and an extra microstate E appeared in N3 stage.
微观状态反映了大脑的瞬时状态，可以在很高的时间分辨率下识别大脑整体功能状态的不连续和非线性变化 , 。已经发现四种典型微观状态A、B、C和D与后扣带皮层的活动有关 。布罗贝克等人。研究健康受试者的觉醒和NREM睡眠，发现W、N1和N3阶段微状态C占主导地位，而N2阶段微状态B占主导地位。随着睡眠深度的增加，微状态D的参数GEV逐渐下降 。库恩等人。等人调查了发作性睡病患者的早期 NREM 睡眠，发现 N1 阶段微状态 C 和 D 占主导地位，N2 和 N3 阶段微状态 D 仍占主导地位，N3 阶段出现了额外的微状态 E 。

Through microstate analysis of controls, OSA_OH and OSA_withoutOH patients in 4 sleep stages, we found that the microstate C and D changed during 4 sleep periods in OSA_OH patients compared to controls and OSA_withoutOH patients. The parameter Time Coverage of microstate C increased, while Occurrence of microstate D decreased. Previous studies have shown that microstates C, D were a phenotype of schizophrenia. da Cruz et al. found that patients with schizophrenia and their siblings showed increased presence of microstate class C and decreased presence of microstate class D compared to controls. de Bock et al. found that microstate D was significantly decreased in psychosis in ultra-high-risk (UHR) patients with a future psychotic transition, suggesting its potential as a selective biomarker of future transition in UHR patients. Kuhn et al. found that the duration of all the microstates in the N3 stage of narcoleptic patients was smaller than that of controls, and the authors believed that the microstate D of narcoleptic patients played a more important role than that of controls. Microstate D was associated with attention networks according to EEG-fMRI studies, so Kuhn et al. believed that the persistence of activities in the attention network of narcolepsy patients during sleep was higher. Therefore, we believed that microstates C and D might also be a potential biomarker for OSA patients.
通过对对照组、OSA_OH和OSA_withoutOH患者在4个睡眠阶段的微状态分析，我们发现与对照组和OSA_withoutOH患者相比，OSA_OH患者在4个睡眠阶段的微状态C和D发生了变化。微状态C的时间覆盖率增加，而微状态D的出现次数减少。先前的研究表明，微观状态C、D是精神分裂症的一种表型 。达克鲁兹等人。发现与对照组相比，精神分裂症患者及其兄弟姐妹的 C 类微状态增加，D 类微状态减少 。德博克等人。发现在未来有精神病转变的超高危 (UHR) 患者的精神病中，微状态 D 显着降低，这表明其作为 UHR 患者未来转变的选择性生物标志物的潜力 。库恩等人。研究发现，发作性睡病患者 N3 阶段所有微状态的持续时间均小于对照者，作者认为发作性睡病患者的微状态 D 比对照者发挥了更重要的作用 。根据 EEG-fMRI 研究 ，微状态 D 与注意力网络相关，因此 Kuhn 等人。认为发作性睡病患者在睡眠期间注意力网络活动的持续性较高 。因此，我们认为微状态 C 和 D 也可能是 OSA 患者的潜在生物标志物。

In our study, we also found that the fifth microstate E appeared in N1-OH, but the global variance of microstate E was low, only 9.0%. Although the proportion of microstate E in our study was small, it could not be considered that microstate E was caused by noise. First of all, statistical analysis showed that the parameters of microstate E (Gev (P = 0.000), TC (P = 0.000) and Oc (P = 0.000)) in OSA_OH patients were significantly different. Secondly, CARTOOL software was used to perform K-means clustering for the four sleep periods of OSA_OH and OSA_withoutoh patients. Except for the N1-OH stage, the optimal number of clustering calculated at other sleep stages was 8–15. Topographic map templates obtained from K-means clustering included four typical microstates A, B, C and D, as well as microstate E. In contrast, the optimal number of clusters in the four sleep stages in controls was always 4. When we set the number of clusters to 5–15 and checked these topographic map templates, no microstate E was found in them. Therefore, we thought that the microstate E might exist in both OSA_OH and OSA_withoutOH patients.
在我们的研究中，我们还发现第五个微态E出现在N1-OH中，但微态E的全局方差较低，仅为9.0%。虽然我们研究中微状态E所占的比例较小，但不能认为微状态E是由噪声引起的。首先，统计分析显示OSA_OH患者的微状态E参数（Gev（P = 0.000）、TC（P = 0.000）和Oc（P = 0.000））存在显着差异。其次，使用CARTOOL软件对OSA_OH和OSA_withoutoh患者的四个睡眠时段进行K均值聚类。除N1-OH阶段外，其他睡眠阶段计算的最佳聚类数为8~15。通过 K 均值聚类获得的地形图模板包括四种典型的微状态 A、B、C 和 D，以及微状态 E。相比之下，对照组的四个睡眠阶段的最佳聚类数始终为 4。簇数为5-15，检查这些地形图模板，没有发现微状态E。因此，我们认为微状态E可能同时存在于OSA_OH和OSA_withoutOH患者中。

In the microstate calculation process, K-means clustering is a common practice, but this method has some defects in the microstate modeling,, and it may not be able to find the optimal number of clustering. In addition, there are only 6 EEG channels in ISRUC SLEEP database. Although previous studies have proved that 4 typical microstate topographic maps were not limited by low spatial sampling, this study only focused on 4 typical microstates, and whether the remaining microstates were affected by the number of electrodes remains to be studied. Therefore, we believed that it was necessary to expand the sample size, increase the number of electrodes, and improve the clustering method to further study the cause of the fifth microstate E in OSA patients in N1-OH.
在微观状态计算过程中，K-means聚类是常见的做法，但是该方法在微观状态建模上存在一些 ，并且可能无法找到最优的聚类数量。另外，ISRUC SLEEP数据库中只有6个EEG通道。虽然之前的研究已经证明4种典型的微态地形图不受低空间采样的限制 ，但本研究仅关注4种典型的微态，其余微态是否受到电极数量的影响还有待研究。因此，我们认为有必要扩大样本量、增加电极数量、改进聚类方法来进一步研究N1-OH OSA患者第五微状态E的原因。

Through correlation analysis, only parameters of microstate D in OSA_OH were correlated with delta and alpha power in N1, and the TPB→C and TPE→C was positively correlated with alpha power. Previous studies have shown that there was no conclusive result on the correlation between the four types of EEG microstates and specific power spectrum distribution,. However, Javed et al. believed that the uncertainty of spectral correlation of microstates involved a variety of factors, which could be eliminated by Hilbert spectral analysis. The authors used Hilbert transform to transform EEG signals into delta, theta, alpha, beta, gamma bands, and then calculated the microstates in each sub-band, and the results showed that the band-wise topographies extracted using the proposed method had statistically significant similarity with full band microstates and achieved high percentage for each band in explaining EEG data variance compared to the traditional filtering method. The authors also believed that an average frequency range of 10–15 Hz dominated the formation and the temporal dynamics of microstates. Milz et al. investigated head-surface localization- or source-dependent power effects on the occurrence of the EEG microstate classes, and found that the EEG microstate topography was predominantly determined by intra-cortical sources in the alpha band. Croce et al. investigated EEG microstates associated with intra- and inter-subject alpha variability, and observed an increase in the metrics of microstate B, with the level of intra-subject amplitude alpha oscillations, together with lower coverage of microstate D and a higher frequency of microstate C. Although their study found the relationship between alpha power and microstate metrics, the authors also pointed out that there was no specificity for alpha power. The modulation effect on microstate metrics is not unique to the alpha band. It may be caused by fluctuations in other frequency bands. Wegner et al. found that resting-state EEG microstates were largely determined by alpha frequencies (8–12 Hz) and microstates occur periodically with twice the alpha frequency. Therefore, we believed that the intensity and spatial distribution of alpha band activity in the cortex of OSA patients changed in N1-OH, leading to changes in microstates C and D, which might also be the cause of microstates E.
通过相关分析，只有OSA_OH中微状态D的参数与N1中的delta和α功率相关，TP B→C和TP E→C与α功率正相关。以往的研究表明，四种脑电微状态与特定功率谱分布之间的相关性还没有结论性的结果 、 。然而，贾维德等人。认为微观状态谱相关性的不确定性涉及多种因素，可以通过希尔伯特谱分析来消除 。作者利用希尔伯特变换将脑电信号变换为 delta、theta、alpha、beta、gamma 波段，然后计算每个子波段的微观状态，结果表明使用该方法提取的波段拓扑具有统计显着性与全频带微观状态相似，并且与传统滤波方法相比，每个频带在解释脑电图数据方差方面实现了高百分比 。作者还认为，10-15 Hz 的平均频率范围主导着微观状态的形成和时间动态 。米尔兹等人。研究了头部表面定位或源依赖性功率对 EEG 微状态类别发生的影响，发现 EEG 微状态拓扑主要由 α 波段中的皮质内源决定。克罗齐等人。 研究了与对象内和对象间 α 变异相关的脑电图微状态，并观察到微状态 B 的指标随着对象内振幅 α 振荡水平的增加，以及微状态 D 的覆盖范围较低和微状态 C 的频率较高 .尽管他们的研究发现了阿尔法功率和微观状态指标之间的关系，但作者也指出阿尔法功率没有特异性。对微观状态指标的调制效应并非阿尔法波段所独有。这可能是由其他频带中的波动引起的。韦格纳等人。发现静息态脑电图微状态很大程度上由α频率（8-12 Hz）决定，并且微状态以两倍α频率周期性发生 。因此，我们认为N1-OH中OSA患者皮质α带活动强度和空间分布发生变化，导致微状态C和D的变化，这也可能是微状态E的原因。

Sample entropy is an improved method for measuring the complexity of time series, and it has applications in evaluating the complexity of physiological time series and diagnosing pathological state,,. Zhou et al. found that the sample entropy of sleep apnea syndrome patients was lower than that of controls in each sleep stage. We found that the sample entropy of OSA patints was significantly different from that of controls only in N1 (P = 0.015). Murphy et al. employed sample entropy to calculate the complexity of the microstate sequence over the entire template length in subjects with psychotic disorders. Their results showed that there was no correlation between sequence length of microstates and entropy in psychiatric patients and controls. Our study also showed that the microstate parameters were not correlated with the sample entropy.
样本熵是一种改进的测量时间序列复杂性的方法，在评估生理时间序列的复杂性和诊断病理状态方面有应用 , , 。周等人。研究发现，睡眠呼吸暂停综合征患者在每个睡眠阶段的样本熵均低于对照者 。我们发现 OSA 患者的样本熵仅在 N1 中与对照组显着不同（P = 0.015）。墨菲等人。利用样本熵来计算精神障碍受试者整个模板长度上微状态序列的复杂性 。他们的结果表明，精神病患者和对照者的微观状态序列长度和熵之间没有相关性 。我们的研究还表明，微观状态参数与样本熵不相关。

EEG signal has a long-term correlation of dynamic oscillation characteristics,. Detrended fluctuation analysis (DFA) quantifies the time-domain fluctuation of time series by power-law method, and describes the scaling behavior or long-range correlation of time series with scale index, which is suitable for studying the correlation of long-range power-law functions of various unstable time series. Our study showed that the scale index of OSA patients and controls was 0.5 < α < 1.0, which indicated that there was a long-range power-law continuous correlation of EEG signal (with self-similarity of fractal dimension). The scale index α of OSA patients was higher than controls in four sleep stages, and the scale index α of OSA patients in N1-OH and N3-OH was higher than that of N2-OH and R-OH. D’Rozario et al. found that the DFA of the OSA patients was higher than controls with eyes opening and closing, and the DFA of the two groups with eyes opening was higher than that with eyes closing. Previous studies have shown that the microstate of healthy subjects exhibit scale-free and self-similar dynamic characteristics . Murphy et al. carried out fractal analysis on the microstate of psychiatric patients, and found that the microstate sequence has a long-term time-dependent. However, our study showed that the microstate parameters were not related to FDA.
EEG 信号具有动态振荡特性长期相关性。去趋势波动分析（DFA）通过幂律方法量化时间序列的时域波动，用尺度指标描述时间序列的尺度行为或长程相关性，适合研究长程幂相关性各种不稳定时间序列的律函数。我们的研究表明，OSA患者和对照的量表指数为0.5 < α < 1.0，这表明EEG信号存在长程幂律连续相关性（具有分形维数的自相似性）。 OSA患者在4个睡眠阶段的量表指数α均高于对照组，其中N1-OH和N3-OH的OSA患者量表指数α高于N2-OH和R-OH。德罗萨里奥等人。发现OSA患者的DFA高于睁眼和闭眼的对照组，且两组睁眼的DFA均高于闭眼的 。先前的研究表明，健康受试者的微观状态表现出无标度和自相似的动态特征 。墨菲等人。对精神病患者的微观状态进行了分形分析，发现微观状态序列具有长期的时间依赖性 。然而，我们的研究表明微观状态参数与 FDA 无关。

However, when the five microstates were used as the fitting template, all the transition probabilities in OSA_OH patients in N1 were not correlated with sample entropy and FDA, while the transition probabilities of microstates A → B, A → E and C → E in N2, N3 and R were correlated with FDA. In OSA_withoutOH patients, there was no correlation between the five microstate parameters in the four sleep stages, but TPX→Y in N1 and R stages were correlated with sample entropy, and TPX→Y in N1, N2 and R stages were correlated with FDA. Our study seemed to show that there was a difference in the correlation of transition probability between OSA_OH and OSA_withoutOH patients after increasing the microstate E, but at present, there were more studies on microstate parameter (Global Explained Variance, Mean Duration, Time coverage and Occurrence,–,,,,,–, but few studies on transfer probability ,.
然而，当以5个微状态为拟合模板时，N1中OSA_OH患者的所有转移概率与样本熵和FDA均不相关，而N2中微状态A→B、A→E和C→E的转移概率、N3和R与FDA相关。在OSA_withoutOH患者中，4个睡眠阶段的5个微状态参数之间不存在相关性，但N1和R阶段的TP X→Y与样本熵相关，N1、N2和R阶段的TP X→Y与样本熵相关。美国食品和药物管理局。我们的研究似乎表明增加微状态E后OSA_OH和OSA_withoutOH患者之间的转移概率相关性存在差异，但目前关于微状态参数（全局解释方差、平均持续时间、时间覆盖和发生次数）的研究较多 , – , , , , , – , 但对转移概率的研究很少 , 。

At present, there is no simple and effective EEG biomarker that can reflect the negative impact of OSA on the brain, although D’Rozario et al. have shown that the DFA scale index has the potential as an EEG biomarker of neurobehavioral damage. However, their study only compared DFA and power spectrum, and lacked comparative analysis with other EEG biomarkers (such as sample entropy, microstate, etc.). In addition, they only considered the single scene of simulated driving, and lacked the research on sleep EEG and its prognostic value. In our study, the sleep EEG of OSA patients was analyzed by the microstate method and the correlation analysis with power, sample entropy and DFA was carried out. The results showed that the microstate C increased presence and the microstate D decreased presence in OSA_OH patients. The fifth microstate E appeared during N1-OH, but the probability of other microstates transferring to microstate E was small. The microstate D in OSA_OH patients in N1-OH was correlated with delta, beta and alpha power; the transition probability of the microstate B → C and E → C was correlated with alpha power. In other sleep stages, the microstate parameters were not correlated with power, sample entropy and FDA. These showed that the microstate also had the potential as a biomarker of OSA EEG. Zappasodi el al. investigated prognostic value of EEG microstates in acute stroke, and found that a preserved microstate B in acute phase correlated with a better effective recovery. Therefore, whether there is a correlation between microstate and OSA score and whether it has prognostic value for OSA patients is our next research work.
目前，还没有简单有效的脑电图生物标志物可以反映 OSA 对大脑的负面影响，尽管 D'Rozario 等人。已表明 DFA 量表指数具有作为神经行为损伤的脑电图生物标志物的潜力 。然而，他们的研究仅比较了DFA和功率谱，缺乏与其他脑电生物标志物（如样本熵、微观状态等）的比较分析。此外，他们只考虑了模拟驾驶的单一场景，缺乏对睡眠脑电图及其预后价值的研究。本研究采用微状态法对OSA患者的睡眠脑电图进行分析，并与功率、样本熵和DFA进行相关性分析。结果表明，在 OSA_OH 患者中，微状态 C 增加了存在感，微状态 D 减少了存在感。第五个微状态E出现在N1-OH期间，但其他微状态转移到微状态E的概率很小。 N1-OH 中 OSA_OH 患者的微状态 D 与 delta、beta 和 alpha 功率相关；微观状态B→C和E→C的转变概率与α功率相关。在其他睡眠阶段，微观状态参数与功率、样本熵和 FDA 不相关。这些表明微状态也具有作为 OSA 脑电图生物标志物的潜力。扎帕索迪等人。研究了脑电图微状态在急性中风中的预后价值，发现急性期保留的微状态 B 与更好的有效恢复相关 。因此，微状态与OSA评分是否存在相关性以及对OSA患者是否具有预后价值是我们下一步的研究工作。

Methods 方法

Data sources 数据来源

The polysomnography (PSG) recordings come from an open-access sleep dataset, ISRUC-Sleep (http://sleeptight.isr.uc.pt/ISRUC_Sleep). The data were obtained from healthy subjects, subjects with sleep disorders, and subjects under the effect of sleep medication (i.e., Subgroup_I, Subgroup_II, Subgroup_III), from all-night PSG records, each with duration around 8 h, which were acquired by the Sleep Medicine Centre of the Hospital of Coimbra University (CHUC). All EEG, EOG, and EMG (chin) recordings were performed with a sampling rate of 200 Hz and stored into computer files using the standard EDF data format. The PSG recordings were composed by signals from 19 channels, of which EEG signal had 6 channels (F3, C3, O1, F4, C4 and O2). All recordings were segmented into epochs of 30 s and visually labeled by two experts according to the guidelines of AASM, with the stages: awake (W), NREM (including N1, N2 and N3) and REM (abbreviated as R) sleep. Our dataset came from 30 OSA patients (excluding patients with other complications and taking medications) form Subgroup_I and 10 healthy subjects form Subgroup_III.
多导睡眠图 (PSG) 记录来自开放获取的睡眠数据集 ISRUC-Sleep ( http://sleeptight.isr.uc.pt/ISRUC_Sleep )。数据来自健康受试者、睡眠障碍受试者和睡眠药物作用下的受试者（即亚组_I、亚组_II、亚组_III）的整夜 PSG 记录，每个记录的持续时间约为 8 小时，由科英布拉大学医院睡眠医学中心 (CHUC) 。所有 EEG、EOG 和 EMG（下巴）记录均以 200 Hz 的采样率进行，并使用标准 EDF 数据格式存储到计算机文件中。 PSG记录由19个通道的信号组成，其中EEG信号有6个通道（F3、C3、O1、F4、C4和O2）。所有录音均按 30 秒为一个周期，由两位专家根据 AASM 的指南进行可视化标记，分为清醒（W）、NREM（包括 N1、N2 和 N3）和 REM（缩写为 R）睡眠阶段。我们的数据集来自 Subgroup_I 中的 30 名 OSA 患者（不包括患有其他并发症和正在服用药物的患者）和 Subgroup_III 中的 10 名健康受试者。

EEG data pre-processing 脑电数据预处理

EEG signals were pre-processed with the EEGLAB toolbox for MATLAB, which were re-referenced to the common average reference, high-pass filtered with a 0.1 Hz zero-phase FIR filter, low-pass filtered with a 45 Hz zero-phase FIR filter, and down-sampled to 100 Hz. EEG signals were inspected for artifacts with a procedure based on Independent Components (ICs) using ADJUST plug-in. IC scalp maps and frequency spectra were inspected, and ICs that displayed features indicative of artifacts were removed.
EEG 信号使用 MATLAB 的 EEGLAB 工具箱进行预处理，重新参考公共平均参考，使用 0.1 Hz 零相位 FIR 滤波器进行高通滤波，使用 45 Hz 零相位 FIR 进行低通滤波滤波器，并下采样至 100 Hz。使用 ADJUST 插件通过基于独立组件 (IC) 的程序检查 EEG 信号是否存在伪影。检查了 IC 头皮图和频谱，并去除了显示出伪影特征的 IC 。

For OSA_OH patients, EEG epochs were extracted for each patient when labeled OH (Obstructive Hypopnea) or OA (Obstructive Apnea). For OSA_withoutOH patients, EEG epochs were extracted for each patient when labeled normal. For healthy controls, epochs were extracted from each healthy subject in each sleep stage (W, N1, N2, N3 and R). For all data, an epoch lasted for 30 s.
对于 OSA_OH 患者，当标记为 OH（阻塞性呼吸暂停）或 OA（阻塞性呼吸暂停）时，为每位患者提取 EEG 时期。对于 OSA_withoutOH 患者，在标记为正常时提取每位患者的 EEG 时期。对于健康对照，从每个健康受试者的每个睡眠阶段（W、N1、N2、N3 和 R）提取时期。对于所有数据，一个 epoch 持续 30 秒。

Microstate analysis 微观状态分析

Microstates reflect the instantaneous state of the brain, and can identify global functional brain states at very high temporal resolution. EEG microstates were extracted from each subject with the CARTOOL software by using a polarity-insensitive K-means algorithm in each epoch. The optimal number of microstates was determined by means of a combination of cross-validation and the Krzanovski–Lai criteria. The same number of microstates was retained for each subject. The microstate maps of each subject were then submitted to a second cluster analysis in order to identify the dominant maps across the subjects. Statistical smoothing was applied to remove temporally isolated topographic maps with low explanatory power. Clusters that correlated above 90% were merged, and segments shorter than 10 ms were rejected. The reference maps were selected as those that highly spatially correlated with the other maps in the same cluster. The microstate maps of each subject were matched with the reference maps showing the higher spatial correlation.
微观状态反映了大脑的瞬时状态，并且可以以非常高的时间分辨率识别大脑的整体功能状态。使用 CARTOOL 软件在每个时期使用极性不敏感的 K 均值算法从每个受试者中提取 EEG 微观状态。最佳微观状态数是通过交叉验证和 Krzanovski-Lai 标准相结合来确定的 。为每个受试者保留相同数量的微观状态。然后将每个受试者的微观状态图提交给第二次聚类分析，以便识别受试者中的主要图 。应用统计平滑来去除解释能力低的时间上孤立的地形图。相关性高于 90% 的簇被合并，短于 10 毫秒的片段被拒绝。选择与同一簇中的其他地图在空间上高度相关的参考地图。每个受试者的微观状态图与参考图相匹配，显示出更高的空间相关性。

The calculated microstate parameters include: Global Explained Variance (Gev (%)), Mean Duration (MD (ms)), Time Coverage (TC (%)), Occurrence (Oc (/s)), Transition Probability (TP).
计算的微观状态参数包括：全局解释方差 (Gev (%))、平均持续时间 (MD (ms))、时间覆盖 (TC (%))、发生率 (Oc (/s))、转移概率 (TP)。

Power spectral analysis 功率谱分析

Previous research has shown that power spectrum analysis of wakefulness EEG is helpful to detect human's alertness–. Greneche et al. have compared the power spectrum of wakefulness EEG between OSA patients and healthy controls, but we focused on the power spectrum of sleep EEG between OSA patients and healthy controls. After artefactual epochs were rejected, power spectrum was obtained using a standard fast Fourier transform (FFT) with a rectangular weighting window, for each non-overlapping 5 s epoch of EEG, i.e., 500 data points ,. Absolute power spectra was calculated in the delta, theta, alpha, sigma and beta bands in each frequency ranges of 0.5–4.5, 4.5–8, 8–12, 12–15 and 15–32 Hz. Power spectrum in each sleep-staged 30 s epoch was calculated by averaging data from 6 5 s epochs. Absolute power spectrum was used to calculate to power density. For example, delta power density is equal to absolute power in the 0.5–4.5 Hz frequency range divided by the sum of absolute powers in 0.5–32 Hz frequency ranges.
先前的研究表明，觉醒脑电图的功率谱分析有助于检测人类的警觉性 – 。格内什等人。比较了 OSA 患者和健康对照者之间的清醒 EEG 功率谱 ，但我们重点关注 OSA 患者和健康对照者之间的睡眠 EEG 功率谱。在拒绝人工纪元之后，对于EEG的每个非重叠5秒纪元，即500个数据点 使用具有矩形加权窗口的标准快速傅里叶变换（FFT）来获得功率谱。计算 0.5–4.5、4.5–8、8–12、12–15 和 15–32 Hz 各频率范围内 delta、theta、alpha、sigma 和 beta 频段的绝对功率谱。每个睡眠阶段的 30 秒时期的功率谱是通过对 6 个 5 秒时期的数据进行平均来计算的。使用绝对功率谱来计算功率密度。例如，功率密度增量等于 0.5–4.5 Hz 频率范围内的绝对功率除以 0.5–32 Hz 频率范围内的绝对功率之和。

Sample entropy analysis 样本熵分析

Sample entropy is a method to measure the complexity of time series, which has been successfully applied in the analysis of physiological signals, such as heart rate, blood pressure, EEG, etc. Its calculation results are related to the selection of parameters m, r and n. (1) The embedding dimension m represents the length of the sequence. Generally, m is set to 1 or 2, because when m > 2, the amount of data n is required to be more than several thousand points. (2) The physical meaning of threshold r is the radius of super ball with dimension m, which is a parameter to measure the similarity of time series, which can be set according to the needs of specific problems. Pincus believed that when r was set to (0.1–0.25) × SD (SD was the standard deviation), and the effective statistical characteristics could be obtained. (3) The input data point n is set to 100–5000 in order to get effective statistical characteristics and small pseudo error for the given data. Therefore, in our study, we took m = 2, r = 0.2SD and n = 1000.
样本熵是一种衡量时间序列复杂程度的方法，已成功应用于生理信号的分析，如心率、血压、脑电图等，其计算结果与参数m 、 r的选择有关和n 。 (1)嵌入维数m表示序列的长度。一般情况下， m设置为1或2，因为当m>2时，要求数据量n大于几千个点 。 (2)阈值r的物理意义是维度为m的超级球的半径，是衡量时间序列相似度的参数，可以根据具体问题的需要进行设置。 Pincus认为，当r设置为（0.1-0.25）×SD（SD为标准差）时，可以获得有效的统计特征 。 (3) 输入数据点n设置为100-5000，以便给定数据获得有效的统计特性和较小的伪误差。因此，在我们的研究中，我们取m = 2、 r = 0.2SD 和n = 1000。

Detrended fluctuation analysis
去趋势波动分析

Detrended fluctuation analysis (DFA) is widely used to analyze the long-range correlation of various unsteady signals, such as ECG, EEG, DNA sequence, weather signal, turbulence velocity and temperature field. DFA is an improved root mean square analysis method, which has two advantages over the commonly used fractal analysis methods: (1) it can detect the self-similarity of time series signal that seems to be unstable but is inherently self-similarity; (2) it can avoid the obvious self-similarity trend caused by external factors.
去趋势波动分析（DFA）广泛用于分析各种非定常信号的长程相关性，例如心电图、脑电图、DNA序列、天气信号、湍流速度和温度场。 DFA是一种改进的均方根分析方法，与常用的分形分析方法相比，它具有两个优点：（1）它可以检测看似不稳定但本质上自相似的时间序列信号的自相似性； (2)可以避免因外界因素造成的明显的自相似趋势 。

The function relationship curve of the DFA wave function F(s) and the interval length s is drawn in double logarithmic coordinates, then the slope of the curve by linear fitting is calculated, which is the scale index α. The scale index α provides a quantitative index for the correlation of the long-range power function. If α < 0.5, it means that the segmented time series are independent of each other; if 0.5 < α < 1.0, it means that the segmented time series have continuous correlation in the form of long-range power rate (with self-similarity of fractal dimension); if α = 1, it indicates that the segmented time series fluctuate in the form of 1/f noise; if 1.0 < α < 1.5, it means that the segmented time series do not have long-range correlation; if α = 1.5, it indicates that the segmented time series are Brownian noise, that is, they are random independent. In our study, we took the data length n = 3000, and divided the sequence into 40 non-overlapping segments.
在双对数坐标下绘制DFA波函数F ( s )与间隔长度s的函数关系曲线，然后通过线性拟合计算该曲线的斜率，即尺度指数α。尺度指标α为长程幂函数的相关性提供了定量指标。如果α < 0.5，则表示分割后的时间序列相互独立；如果0.5 < α < 1.0，则表示分段时间序列以长程功率率的形式具有连续相关性（具有分形维数的自相似性）；如果α=1，则表明分段时间序列以1 / f噪声的形式波动；如果1.0 < α < 1.5，则表示分段时间序列不具有长程相关性；如果α = 1.5，则表明分段时间序列是布朗噪声，即它们是随机独立的。在我们的研究中，我们取数据长度n = 3000，并将序列分为40个不重叠的段 。

Statistical analysis 统计分析

Multivariate ANOVA was performed for the microstate parameters, power, sample entropy and FDA of OSA_OH vs. Control and OSA_OH vs. OSA_withoutOH at each sleep stage, and Bonferroni post-hoc tests were performed. For microstate analysis, 2 × 4 × 4 and 2 × 5 × 4 multivariate ANOVA designs have been performance (2 levels: OSA_OH vs. Control or OSA_OH vs. OSA_ withoutOH), microstate (4 or 5) and Microstate parameters (Gev, MD, TC, Oc)). A two-sided one-sample t test was then performed within groups. Pearson correlation test was used to verify the correlation between microstate parameters and power, sample entropy and FDA. A threshold for significance was assessed at P < 0.00714 (i.e. 0.05/7, 7 being the number of microstates and power, sample entropy and FDA). A percentile-based bootstrap, with 1000 replicate samples, was applied to assess the 95% confidence interval of Pearson’s r values. All statistical analyses were performed using IBM SPSS statistics, version 21 (IBM Corp., Armonk, NY, USA), and P < 0.05 was considered statistically significant. All plots were performed using Matlab R2013b (MathWorks. Inc, USA) and CARTOOL software.
对每个睡眠阶段的 OSA_OH 与对照以及 OSA_OH 与 OSA_withoutOH 的微观状态参数、功率、样本熵和 FDA 进行多变量方差分析，并进行 Bonferroni 事后测试。对于微观状态分析，2 × 4 × 4 和 2 × 5 × 4 多元方差分析设计已获得性能（2 个水平：OSA_OH 与对照或 OSA_OH 与 OSA_ withoutOH）、微观状态（4 或 5）和微观状态参数（Gev、MD） ，TC，OC））。然后在组内进行双边单样本 t 检验。采用Pearson相关性检验验证微观状态参数与功效、样本熵和FDA之间的相关性。显着性阈值评估为 P < 0.00714（即 0.05/7，7 是微观状态和功率、样本熵和 FDA 的数量）。应用基于百分位的引导程序（具有 1000 个重复样本）来评估 Pearson r 值的 95% 置信区间。所有统计分析均使用 IBM SPSS 统计第 21 版（IBM Corp.，Armonk，NY，USA）进行，P < 0.05 被认为具有统计显着性。所有绘图均使用 Matlab R2013b（MathWorks.Inc，美国）和 CARTOOL 软件进行。

Acknowledgements 致谢

This study was funded by Grant nos. 82060329, 11265007 from National Natural Science Foundation of China (NSFC), No. 2020J0052 from Scientific Research Fund Project of Yunnan Education Department of China.
这项研究由拨款号资助。国家自然科学基金项目82060329、11265007，云南省教育厅科研基金项目2020J0052。

Author contributions 作者贡献

X.X. wrote the main manuscript text. Y.R. prepared Tables ​Tables22 and ​and3.3. Shenghan Gao prepared Figure ​Figure3.3. J.H. carried out the overall algorithm design and experiments. All authors reviewed the manuscript.
XX 撰写了主要稿件文本。 YR 准备了表​表2 2和​和3。 3 .高胜汉准备了图​图3。 3 . JH进行了整体算法设计和实验。所有作者都审阅了手稿。

Competing interests 利益争夺

The authors declare no competing interests.
作者声明没有竞争利益。

Footnotes 脚注

The original online version of this Article was revised: The original version of this Article contained an error in Reference 34, which was incorrectly given as: Fvwa, B. et al. EEG microstate periodicity explained by rotating phase patterns of resting-state alpha oscillations. Neuroimage 2, 224 (2020). The correct reference is listed as: von Wegner, F. et al. EEG microstate periodicity explained by rotating phase patterns of resting-state alpha oscillations. Neuroimage 224, 117372 (2021).
本文原网络版已修改：本文原版在参考文献34中有一个错误，错误地给出为：Fvwa, B. et al. EEG 微状态周期性通过静息态 α 振荡的旋转相位模式来解释。神经影像2 , 224 (2020)。正确的参考文献列出为：von Wegner, F. et al。 EEG 微状态周期性通过静息态 α 振荡的旋转相位模式来解释。神经影像224 , 117372 (2021)。

Publisher's note 出版商备注

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
施普林格·自然对于已出版的地图和机构隶属关系中的管辖权主张保持中立。

Change history 变更历史记录

10/21/2021

A Correction to this paper has been published: 10.1038/s41598-021-00538-6 IF: 3.8 Q1
对本文的更正已发布：10.1038/s41598-021-00538-6IF：3.8 Q1

References 参考

1. Heinzer R, Vat S, Marques-Vidal P, Marti-Soler H, Andries D, Tobback N, et al. Prevalence of sleep-disordered breathing in the general population: The HypnoLaus study. Lancet Respir. Med. 2015;3(4):310–318. doi: 10.1016/S2213-2600(15)00043-0 IF: 38.7 Q1 . [PMC free article] [PubMed] [CrossRef] []
1. Heinzer R、Vat S、Marques-Vidal P、Marti-Soler H、Andries D、Tobback N 等人。一般人群中睡眠呼吸障碍的患病率：HypnoLaus 研究。柳叶刀呼吸。医学。 2015年； 3 （4）：310-318。号码：10.1016/S2213-2600(15)00043-0 如果：38.7 Q1 。 [ PMC 免费文章] [ PubMed ] [ CrossRef ] [ ]
2. Young T, Peppard P, Palta M, Hla K, Finn L, Morgan B, et al. Population-based study of sleep-disordered breathing as a risk factor for hypertension. Arch. Intern Med. 1997;157(15):1746–1752. doi: 10.1001/archinte.1997.00440360178019 . [PubMed] [CrossRef] []
2. Young T、Peppard P、Palta M、Hla K、Finn L、Morgan B 等人。基于人群的睡眠呼吸障碍作为高血压危险因素的研究。拱。实习医生。 1997 年； 157 （15）：1746-1752。 doi: 10.1001/archinte.1997.00440360178019 。 [公共医学] [交叉引用] []
3. Eckert DJ, Malhotra A. Pathophysiology of adult obstructive sleep apnea. Proc. Am. Thorac. Soc. 2008;5(2):144–153. doi: 10.1513/pats.200707-114MG . [PMC free article] [PubMed] [CrossRef] []
3. Eckert DJ，Malhotra A。成人阻塞性睡眠呼吸暂停的病理生理学。过程。是。胸廓。苏克。 2008年； 5 （2）：144-153。 doi: 10.1513/pats.200707-114MG 。 [ PMC 免费文章] [ PubMed ] [ CrossRef ] [ ]
4. George CF. Sleep apnea, alertness, and motor vehicle crashes. Am. J. Respir. Crit. Care Med. 2007;176:954–956. doi: 10.1164/rccm.200605-629PP IF: 19.3 Q1 . [PubMed] [CrossRef] []
4.乔治·CF。睡眠呼吸暂停、警觉性和机动车事故。是。 J.呼吸。暴击。护理医学。 2007年； 176 ：954–956。编号：10.1164/rccm.200605-629PP 如果：19.3 Q1 。 [公共医学] [交叉引用] []
5. Aksahin M, Aydin S, Firat H, Erogul O. Artificial apnea classification with quantitative sleep EEG synchronization. J. Med. Syst. 2012;36(1):139–144. doi: 10.1007/s10916-010-9453-8 IF: 3.5 Q1 . [PubMed] [CrossRef] []
5. Aksahin M、Aydin S、Firat H、Erogul O。人工呼吸暂停分类与定量睡眠脑电图同步。 J. Med。系统。 2012年； 36 （1）：139-144。号码：10.1007/s10916-010-9453-8 如果：3.5 Q1 。 [公共医学] [交叉引用] []
6. Liu D, Pang Z, Lloyd SR. Aneural network method for detection of obstructive sleep apnea and narcolepsy based on pupil size and EEG. IEEE Trans. Neural Netw. 2008;19(2):308–318. doi: 10.1109/TNN.2007.908634 . [PubMed] [CrossRef] []
6.刘D，庞Z，劳埃德SR。基于瞳孔大小和脑电图检测阻塞性睡眠呼吸暂停和发作性睡病的神经网络方法。 IEEE 传输。神经网络。 2008年； 19 （2）：308-318。号码：10.1109/TNN.2007.908634 。 [公共医学] [交叉引用] []
7. Zhou J, Wu XM. Electroencephalogram of sleep apnea syndrome based on sample entropy. Chin. J. Med. Phys. 2016;7:722–725. []
7.周杰，吴晓明。基于样本熵的睡眠呼吸暂停综合征脑电图。下巴。 J. Med。物理。 2016年； 7 ：722–725。 []
8. Grenèche J, Krieger J, Erhardt C, Bonnefond A, Eschenlauer A, Muzet A, et al. EEG spectral power and sleepiness during 24 h of sustained wakefulness in patients with obstructive sleep apnea syndrome. Clin. Neurophysiol. 2008;119:418–428. doi: 10.1016/j.clinph.2007.11.002 IF: 3.7 Q1 . [PubMed] [CrossRef] []
8. Grenèche J、Krieger J、Erhardt C、Bonnefond A、Eschenlauer A、Muzet A 等人。阻塞性睡眠呼吸暂停综合征患者 24 小时持续清醒期间的 EEG 频谱功率和困倦。临床。神经生理学。 2008年； 119 ：418–428。 doi：10.1016/j.clinph.2007.11.002 如果：3.7 Q1 。 [公共医学] [交叉引用] []
9. D’Rozario AL, Kim JW, Wong K. A new EEG biomarker of neurobehavioural impairment and sleepiness in sleep apnea patients and controls during extended wakefulness. Clin. Neurophysiol. 2013;124(8):1605–1614. doi: 10.1016/j.clinph.2013.02.022 IF: 3.7 Q1 . [PubMed] [CrossRef] []
9. D'Rozario AL、Kim JW、Wong K。睡眠呼吸暂停患者和对照组在长时间清醒状态下的神经行为损伤和嗜睡的新脑电图生物标志物。临床。神经生理学。 2013年； 124 （8）：1605-1614。 doi：10.1016/j.clinph.2013.02.022 如果：3.7 Q1 。 [公共医学] [交叉引用] []
10. Vakulin A, D’Rozario A, Kim JW. Quantitative sleep EEG and polysomnographic predictors of driving simulator performance in obstructive sleep apnea. Clin. Neurophysiol. 2015;127:1428–1435. doi: 10.1016/j.clinph.2015.09.004 IF: 3.7 Q1 . [PubMed] [CrossRef] []
10. Vakulin A、D'Rozario A、Kim JW。阻塞性睡眠呼吸暂停驾驶模拟器性能的定量睡眠脑电图和多导睡眠图预测因子。临床。神经生理学。 2015年； 127 ：1428–1435。 doi：10.1016/j.clinph.2015.09.004如果：3.7 Q1 。 [公共医学] [交叉引用] []
11. Jong WK, Shin HB, Robinson PA. Quantitative study of the sleep onset period via detrended fluctuation analysis: Normal vs. narcoleptic subjects. Clin. Neurophysiol. 2009;120:1245–1251. doi: 10.1016/j.clinph.2009.04.018 IF: 3.7 Q1 . [PubMed] [CrossRef] []
11. Jong WK、Shin HB、Robinson PA。通过去趋势波动分析对睡眠开始期进行定量研究：正常与发作性睡病受试者。临床。神经生理学。 2009年； 120 ：1245–1251。 doi：10.1016/j.clinph.2009.04.018如果：3.7 Q1 。 [公共医学] [交叉引用] []
12. Lehmann D, Ozaki H, Pal I. EEG alpha map series: Brain micro-states by space-oriented adaptive segmentation. Electroencephalogr. Clin. Neurophysiol. 1987;67:271–288. doi: 10.1016/0013-4694(87)90025-3 . [PubMed] [CrossRef] []
12. Lehmann D、Ozaki H、Pal I。EEG alpha 图系列：通过面向空间的自适应分割实现大脑微观状态。脑电图学家。临床。神经生理学。 1987； 67 ：271–288。号码：10.1016/0013-4694(87)90025-3。 [公共医学] [交叉引用] []
13. Murray MM, Brunet D, Michel CM. Topographic ERP analyses: A step-by-step tutorial review. Brain Topogr. 2008;20:249–264. doi: 10.1007/s10548-008-0054-5 IF: 2.3 Q3 . [PubMed] [CrossRef] []
13.穆雷 MM、布鲁内特 D、米歇尔 CM。地形 ERP 分析：分步教程回顾。脑拓扑图。 2008年； 20 ：249–264。 DOI：10.1007/s10548-008-0054-5如果：2.3 Q3 。 [公共医学] [交叉引用] []
14. Zappasodi F, Croce P, Giordani A, et al. Prognostic value of EEG microstates in acute stroke. Brain Topogr. 2017;30(5):1–13. doi: 10.1007/s10548-017-0572-0 IF: 2.3 Q3 . [PubMed] [CrossRef] []
14. Zappasodi F、Croce P、Giordani A 等人。脑电图微状态对急性卒中的预后价值。脑拓扑图。 2017年； 30 （5）：1-13。号码：10.1007/s10548-017-0572-0如果：2.3 Q3 。 [公共医学] [交叉引用] []
15. Khanna A, Pascualleone A, Michel CM, Farzan F. Microstates in resting-state EEG: Current status and future directions. Neurosci. Biobehav. Rev. 2015;49:105–113. doi: 10.1016/j.neubiorev.2014.12.010 IF: 7.5 Q1 . [PMC free article] [PubMed] [CrossRef] []
15. Khanna A、Pascualleone A、Michel CM、Farzan F。静息态脑电图的微观状态：现状和未来方向。神经科学。生物行为。 2015 年修订； 49 ：105–113。 doi：10.1016/j.neubiorev.2014.12.010如果：7.5 Q1 。 [ PMC 免费文章] [ PubMed ] [ CrossRef ] [ ]
16. Brodbeck V, Kuhn A, Wegner FV, Morzelewski A. EEG microstates of wakefulness and NREM sleep. Neuroimage. 2012;62(3):2129–2139. doi: 10.1016/j.neuroimage.2012.05.060 IF: 4.7 Q1 . [PubMed] [CrossRef] []
16. Brodbeck V、Kuhn A、Wegner FV、Morzelewski A。觉醒和 NREM 睡眠的脑电图微观状态。神经影像。 2012；62(3)：2129–2139。 doi：10.1016/j.neuroimage.2012.05.060IF：4.7 Q1。 [ PubMed ] [ 交叉引用 ] [ 谷歌学术 ]
17. Kuhn A, Brodbeck V, Tagliazucchi E, Morzelewski A. Narcoleptic patients show fragmented EEG-microstructure during early NREM sleep. Brain Topogr. 2014;28(4):619–635. doi: 10.1007/s10548-014-0387-1 IF: 2.3 Q3 . [PubMed] [CrossRef] []
17. Kuhn A、Brodbeck V、Tagliazucchi E、Morzelewski A。发作性睡病患者在早期 NREM 睡眠期间表现出碎片化的脑电图微结构。脑拓扑图。 2014年； 28 （4）：619–635。号码：10.1007/s10548-014-0387-1如果：2.3 Q3 。 [公共医学] [交叉引用] []
18. Murphy M, Stickgold R, Öngür D. Electroencephalogram microstate abnormalities in early-course psychosis. Biol. Psychiatry Cogn. Neurosci. Neuroimaging. 2020;5(1):35–44. [PMC free article] [PubMed] []
18. Murphy M、Stickgold R、Öngür D。早期精神病中的脑电图微状态异常。生物。精神病学认知。神经科学。神经影像学。 2020； 5 （1）：35-44。 [ PMC 免费文章] [ PubMed ] [ ]
19. Brunet D, Murray MM, Michel CM. Spatiotemporal analysis of multichannel EEG: Cartool. Comput. Intell. Neurosci. 2011;20:813–870. [PMC free article] [PubMed] []
19.布鲁内特·D、默里·MM、米歇尔·CM。多通道脑电图时空分析：Cartool。计算。英特尔。神经科学。 2011； 20 ：813–870。 [ PMC 免费文章] [ PubMed ] [ ]
20. Pascqual-Marqui RD, Michel CM, Lehmann D. Segmnetation of brain electrical activity into microstates: Model estimation and validation. IEEE Trans. Biomed. Eng. 1995;42:658–665. doi: 10.1109/10.391164 IF: 4.4 Q2 . [PubMed] [CrossRef] []
20. Pascqual-Marqui RD、Michel CM、Lehmann D。将脑电活动分割为微观状态：模型估计和验证。 IEEE 传输。生物医学。工程师。 1995； 42 ：658–665。号码：10.1109/10.391164如果：4.4 Q2 。 [公共医学] [交叉引用] []
21. Pascual-Marqui RD, Lehmann D, Faber P, Milz P, Kochi K, Yoshimura M, et al. The resting microstate networks (RMN): Cortical distributions, dynamics, and frequency specific information flow. Quant. Biol. 2014;20:14. []
21. Pascual-Marqui RD、Lehmann D、Faber P、Milz P、Kochi K、Yoshimura M 等。静息微状态网络 (RMN)：皮质分布、动态和频率特定信息流。定量。生物。 2014年； 20:14 。 []
22. Rieger K, Diaz Hernandez L, Baenninger A, Koenig T. 15 years of microstate research in schizophrenia—where are we? A meta-analysis. Front. Psychiatry. 2016;7:22. doi: 10.3389/fpsyt.2016.00022 IF: 3.2 Q2 . [PMC free article] [PubMed] [CrossRef] []
22. Rieger K、Diaz Hernandez L、Baenninger A、Koenig T。15 年精神分裂症微观状态研究——我们在哪里？荟萃分析。正面。精神病学。 2016年； 7:22 。 doi: 10.3389/fpsyt.2016.00022 IF: 3.2 Q2 如果：3.2 Q2 。 [ PMC 免费文章] [ PubMed ] [ CrossRef ] [ ]
23. Cruz J, Favrod O, Roinishvili M, et al. EEG microstates are a candidate endophenotype for schizophrenia. Nat. Commun. 2020;11(1):1. doi: 10.1038/s41467-019-13993-7 IF: 14.7 Q1 . [PMC free article] [PubMed] [CrossRef] []
23. Cruz J、Favrod O、Roinishvili M 等。脑电图微状态是精神分裂症的候选内表型。纳特。交流。 2020； 11 （1）：1。号码：10.1038/s41467-019-13993-7 如果：14.7 Q1 。 [ PMC 免费文章] [ PubMed ] [ CrossRef ] [ ]
24. Bock RD, Mackintosh AJ, Maier F, et al. EEG microstates as biomarker for psychosis in ultra-high-risk patients. Transl. Psychiatry. 2020;10(1):300. doi: 10.1038/s41398-020-00963-7 IF: 5.8 Q1 . [PMC free article] [PubMed] [CrossRef] []
24. Bock RD、Mackintosh AJ、Maier F 等人。脑电图微状态作为超高危患者精神病的生物标志物。译。精神病学。 2020； 10 （1）：300。号码：10.1038/s41398-020-00963-7 如果：5.8 Q1 。 [ PMC 免费文章] [ PubMed ] [ CrossRef ] [ ]
25. Britz J, Van De Ville D, Michel CM. BOLD correlates of EEG topography reveal rapid resting-state network dynamics. Neuroimage. 2010;52:1162–1170. doi: 10.1016/j.neuroimage.2010.02.052 IF: 4.7 Q1 . [PubMed] [CrossRef] []
25.布里茨·J，范·德维尔·D，米歇尔·CM。脑电图地形的 BOLD 相关性揭示了快速的静息态网络动态。神经影像。 2010； 52 ：1162–1170。 doi：10.1016/j.neuroimage.2010.02.052 如果：4.7 Q1 。 [公共医学] [交叉引用] []
26. Dinov M, Leech R. Modeling uncertainties in EEG microstates: Analysis of real and imagined motor movements using probabilistic clustering-driven training of probabilistic neural networks. Front. Hum. Neurosci. 2017;11:534. doi: 10.3389/fnhum.2017.00534 IF: 2.4 Q2 . [PMC free article] [PubMed] [CrossRef] []
26. Dinov M，Leech R。脑电图微观状态的不确定性建模：使用概率聚类驱动的概率神经网络训练分析真实和想象的运动运动。正面。哼。神经科学。 2017年； 11 :534。 doi: 10.3389/fnhum.2017.00534 IF: 2.4 Q2 如果：2.4 Q2 。 [ PMC 免费文章] [ PubMed ] [ CrossRef ] [ ]
27. Michel CM, Koenig T. EEG microstates as a tool for studying the temporal dynamics of whole-brain neuronal networks: A review. Neuroimage. 2018;180(pt B):577–593. doi: 10.1016/j.neuroimage.2017.11.062 IF: 4.7 Q1 . [PubMed] [CrossRef] []
27. Michel CM，Koenig T。脑电图微状态作为研究全脑神经元网络时间动态的工具：综述。神经影像。 2018； 180 （B 部分）：577–593。 doi：10.1016/j.neuroimage.2017.11.062 如果：4.7 Q1 。 [公共医学] [交叉引用] []
28. Khalighi S, Sousa T, Santos J, Nunes U. ISRUC-Sleep: A comprehensive public dataset for sleep researchers. Comput. Methods Programs Biomed. 2016;124:180–192. doi: 10.1016/j.cmpb.2015.10.013 IF: 4.9 Q1 . [PubMed] [CrossRef] []
28. Khalighi S、Sousa T、Santos J、Nunes U. ISRUC-Sleep：睡眠研究人员的综合公共数据集。计算。方法计划生物医学。 2016年； 124 ：180–192。 doi：10.1016/j.cmpb.2015.10.013 如果：4.9 Q1 。 [公共医学] [交叉引用] []
29. Khanna A, Pascualleone A, Farzan F. Reliability of resting-state microstate features in electroencephalography. PLoS One. 2014;12:1–12. [PMC free article] [PubMed] []
29. Khanna A、Pascualleone A、Farzan F。脑电图静息态微状态特征的可靠性。公共科学图书馆一号。 2014年； 12 ：1-12。 [ PMC 免费文章] [ PubMed ] [ ]
30. Musso F, Brinkmeyer J, Mobascher A, Warbrick T, Winterer G. Spontaneous brain activity and EEG microstates. A novel EEG/fMRI analysis approach to explore resting-state networks. Neuroimage. 2010;52:1149–1161. doi: 10.1016/j.neuroimage.2010.01.093 IF: 4.7 Q1 . [PubMed] [CrossRef] []
30. Musso F、Brinkmeyer J、Mobascher A、Warbrick T、Winterer G。自发脑活动和脑电图微观状态。一种探索静息态网络的新型脑电图/功能磁共振成像分析方法。神经影像。 2010； 52 ：1149–1161。 doi：10.1016/j.neuroimage.2010.01.093 如果：4.7 Q1 。 [公共医学] [交叉引用] []
31. Ehtasham, Javed, Pierpaolo, et al. Hilbert spectral analysis of EEG data reveals spectral dynamics associated with microstates-ScienceDirect. J. Neurosci. Methods. 2019;325(2):108317–108317. [PubMed] []
31. Ehtasham、Javed、Pierpaolo 等人。脑电图数据的希尔伯特谱分析揭示了与微观状态相关的谱动态 - ScienceDirect。 J.神经科学。方法。 2019； 325 （2）：108317–108317。 [考研] []
32. Milz P, Pascual-Marqui RD, Achermann P, Kochi K, Faber PL. The EEG microstate topography is predominantly determined by intracortical sources in the alpha band. Neuroimage. 2017;162:353–361. doi: 10.1016/j.neuroimage.2017.08.058 IF: 4.7 Q1 . [PubMed] [CrossRef] []
32. Milz P、Pascual-Marqui RD、Achermann P、Kochi K、Faber PL。 EEG 微状态地形主要由 α 波段的皮质内源决定。神经影像。 2017年； 162 ：353–361。 doi：10.1016/j.neuroimage.2017.08.058 如果：4.7 Q1 。 [公共医学] [交叉引用] []
33. Croce P, Quercia A, Costa S, Zappasodi F. EEG microstates associated with intra- and inter-subject alpha variability. Sci. Rep. 2020;10:2469. doi: 10.1038/s41598-020-58787-w IF: 3.8 Q1 . [PMC free article] [PubMed] [CrossRef] []
33. Croce P、Quercia A、Costa S、Zappasodi F。脑电图微状态与受试者内和受试者间 α 变异性相关。科学。 2020 年代表； 10 :2469。 DOI：10.1038/s41598-020-58787-w 如果：3.8 Q1 。 [ PMC 免费文章] [ PubMed ] [ CrossRef ] [ ]
34. von Wegner F, Sb B, Fr B, et al. EEG microstate periodicity explained by rotating phase patterns of resting-state alpha oscillations. Neuroimage. 2021;224:117372. doi: 10.1016/j.neuroimage.2020.117372 IF: 4.7 Q1 . [PubMed] [CrossRef] []
34. von Wegner F、Sb B、Fr B 等。 EEG 微状态周期性通过静息态 α 振荡的旋转相位模式来解释。神经影像。 2021 年； 224 ：117372。 doi：10.1016/j.neuroimage.2020.117372 如果：4.7 Q1 。 [公共医学] [交叉引用] []
35. Takahashi T. Complexity of spontaneous brain activity in mental disorders. Prog. Neuropsychopharmacol. Biol. Psychiatry. 2013;45:258–266. doi: 10.1016/j.pnpbp.2012.05.001 IF: 5.3 Q1 . [PubMed] [CrossRef] []
35. Takahashi T. 精神障碍中自发大脑活动的复杂性。程序。神经精神药理学。生物。精神病学。 2013年； 45 ：258–266。 doi：10.1016/j.pnpbp.2012.05.001 如果：5.3 Q1 。 [公共医学] [交叉引用] []
36. Fernández A, Gómez C, Hornero R, López-Ibor JJ. Complexity and schizophrenia. Prog. Neuropsychopharmacol. Biol. Psychiatry. 2013;45:267–276. doi: 10.1016/j.pnpbp.2012.03.015 IF: 5.3 Q1 . [PubMed] [CrossRef] []
36.费尔南德斯 A、戈麦斯 C、霍内罗 R、洛佩斯-伊博尔 JJ。复杂性和精神分裂症。程序。神经精神药理学。生物。精神病学。 2013年； 45 ：267–276。 doi：10.1016/j.pnpbp.2012.03.015 如果：5.3 Q1 。 [公共医学] [交叉引用] []
37. Watters PA, Martin F. A method for establishing long-range power law correlation from the electroencephalogram. Biol. Psychiatry. 2004;66:79–89. doi: 10.1016/j.biopsycho.2003.09.001 IF: 2.7 Q1 . [PubMed] [CrossRef] []
37. Watters PA，Martin F。一种根据脑电图建立远程幂律相关性的方法。生物。精神病学。 2004年； 66 ：79–89。 doi：10.1016/j.biopsycho.2003.09.001 如果：2.7 Q1 。 [公共医学] [交叉引用] []
38. Nikulin VV, Brismar T. Long-range temporal correlation in electroencephalographic oscillations: Relation to topography, frequency band, age and gender. Neuroscience. 2005;130:549–558. doi: 10.1016/j.neuroscience.2004.10.007 IF: 2.9 Q2 . [PubMed] [CrossRef] []
38. Nikulin VV，Brismar T。脑电图振荡的长程时间相关性：与地形、频段、年龄和性别的关系。神经科学。 2005年； 130 ：549–558。 doi：10.1016/j.neuroscience.2004.10.007 如果：2.9 Q2 。 [公共医学] [交叉引用] []
39. Ville DVD, Britz J, Michel CM. EEG microstate sequences in healthy humans at rest reveal scale-free dynamics. Proc. Natl. Acad. Sci. 2010;107(42):18179–18184. doi: 10.1073/pnas.1007841107 IF: 9.4 Q1 . [PMC free article] [PubMed] [CrossRef] []
39.Ville DVD，布里茨·J，米歇尔·CM。健康人静息时的脑电图微状态序列揭示了无标度动态。过程。国家。阿卡德。科学。 2010； 107 （42）：18179-18184。 DOI：10.1073/pnas.1007841107 如果：9.4 Q1 。 [ PMC 免费文章] [ PubMed ] [ CrossRef ] [ ]
40. Iber, C., Ancoli-Israel, S., Chesson, A., & Quan, S. The AASM manual for the scoring of sleep and associated events: Rules terminology and technical specifications, (2007).
40. Iber, C.、Ancoli-Israel, S.、Chesson, A. 和 Quan, S. AASM 睡眠及相关事件评分手册：规则术语和技术规范，（2007 年）。
41. Brunner C, Delorme A, Makeig S. EEGLAB—an open source MATLAB toolbox for electrophysiological research. Biomed. Eng. 2013;58:3234. [PubMed] []
41. Brunner C、Delorme A、Makeig S. EEGLAB — 用于电生理学研究的开源 MATLAB 工具箱。生物医学。工程师。 2013年； 58 ：3234。 [考研] []
42. Mognon A, Jovicich J, Bruzzone L, Buiatti M. ADJUST: An automatic EEG artifact detector based on the joint use of spatial and temporal features. Psychophysiology. 2011;2:229–240. doi: 10.1111/j.1469-8986.2010.01061.x IF: 2.9 Q1 . [PubMed] [CrossRef] []
42. Mognon A，Jovicich J，Bruzzone L，Buiatti M.ADJUST：基于空间和时间特征联合使用的自动脑电图伪影检测器。心理生理学。 2011； 2 ：229–240。 doi：10.1111/j.1469-8986.2010.01061.x 如果：2.9 Q1 。 [公共医学] [交叉引用] []
43. Minguillon J, Pirondini E, Coscia M, Leeb R. Modular Organization of Reaching and Grasping Movements Investigated Using EEG Microstates. Engineering in Medicine & Biology Society; 2014. [PubMed] []
43. Minguillon J、Pirondini E、Coscia M、Leeb R。使用脑电图微状态研究的伸手和抓取运动的模块化组织。医学与生物学工程学会； 2014. [考研] []
44. Higuchi S, Liu Y, Yuasa T, Maeda A, Motohashi Y. Diurnal variation in alpha power density and subjective sleepiness while performing repeated vigilance tasks. Clin. Neurophysiol. 2001;112:997–1000. doi: 10.1016/S1388-2457(01)00527-2 IF: 3.7 Q1 . [PubMed] [CrossRef] []
44. Higuchi S、Liu Y、Yuasa T、Maeda A、Motohashi Y。执行重复警戒任务时阿尔法功率密度和主观睡意的昼夜变化。临床。神经生理学。 2001年； 112 ：997–1000。号码：10.1016/S1388-2457(01)00527-2 如果：3.7 Q1 。 [公共医学] [交叉引用] []
45. Sforza E, Grandin S, Jouny C, Rochat T, Ibanez V. Is waking electroencephalographic activity a predictor of daytime sleepiness in sleep-related breathing disorders? Eur. Respir. J. 2002;19:645–652. doi: 10.1183/09031936.02.00267302 IF: 16.6 Q1 . [PubMed] [CrossRef] []
45. Sforza E、Grandin S、Jouny C、Rochat T、Ibanez V。清醒时脑电图活动是否是睡眠相关呼吸障碍中白天嗜睡的预测因子？欧元。呼吸。 J.2002 ； 19 ：645–652。号码：10.1183/09031936.02.00267302如果：16.6 Q1 。 [公共医学] [交叉引用] []
46. Wang CW, Guragain B, Verma A, Archer L. Spectral analysis of EEG during microsleep events annotated via driver monitoring system to characterize drowsiness. IEEE Trans. Aerosp. Electron. Syst. 2020;56(2):1346–1356. doi: 10.1109/TAES.2019.2933960 IF: 5.1 Q1 . [CrossRef] []
46. Wang CW、Guragain B、Verma A、Archer L。通过驾驶员监控系统注释微睡眠事件期间的脑电图频谱分析，以表征睡意。 IEEE 传输。气溶胶。电子。系统。 2020； 56 （2）：1346-1356。编号：10.1109/TAES.2019.2933960 如果：5.1 Q1 。 [交叉引用] []
47. Press WH, Teukolsky SA, Vetterling WT, Flannery BP. Numerical Recipe in C. 2. Cambridge University Press; 1992. []
47.按 WH、Teukolsky SA、Vetterling WT、Flannery BP。 C 中的数值配方。 2. 剑桥大学出版社； 1992. []
48. Pincus S. Approximate entropy (ApEn) as a complexity measure. Chaos. 1995;5(1):110–117. doi: 10.1063/1.166092 IF: 2.7 Q1 . [PubMed] [CrossRef] []
48. Pincus S. 近似熵（ApEn）作为复杂性度量。混乱。 1995； 5 （1）：110-117。号码：10.1063/1.166092 如果：2.7 Q1 。 [考研] [交叉引用] []
49. Parish LM, Worrell GA, Cranstoun SD. Long-range temporal correlation in epileptogenic and non-epileptogenic human hippocampus. Neuroscience. 2004;125(4):1069–1076. doi: 10.1016/j.neuroscience.2004.03.002 IF: 2.9 Q2 . [PubMed] [CrossRef] []
49.教区 LM，佐治亚州沃勒尔，南达科他州克兰斯顿。致癫痫和非致癫痫人类海马的长程时间相关性。神经科学。 2004年； 125 （4）：1069–1076。 doi：10.1016/j.neuroscience.2004.03.002 如果：2.9 Q2 。 [公共医学] [交叉引用] []
50. Wang CF, Zhang LX, Liu S, Sun CC, Wang YJ, Zhao X, et al. Feature extraction and recognition of resting EEG in poststroke depression subjects based on detrended fluctuation analysis. Chin. J. Biomed. Eng. 2013;32(5):520–525. []
50.王CF，张LX，刘S，孙CC，王YJ，赵X，等。基于去趋势波动分析的脑卒中后抑郁受试者静息脑电图特征提取和识别下巴。 J.生物医学。工程师。 2013年； 32 （5）：520-525。 []
Articles from Scientific Reports are provided here courtesy of Nature Publishing Group
科学报告中的文章由自然出版集团提供
保存
cn 中文
简体中文
English 英语
用户头像
退出登陆
lonely cat 孤独的猫
PRO 专业版
FREE 自由的
收藏当前网页
截图
上传本地文件
保存至GUGA
(0/0)
升级Pro解锁功能
仅限Pro用户使用，请先解锁Pro。
知道了
notification 通知
保存成功