毕业设计（论文）外文翻译-手机睡眠监测系统的实现与设计

1、1外文资料翻译译文基于智能手机的非接触睡眠呼吸暂停检测Rajalakshmi Nandakumar Shyamnath Gollakota Nathaniel Watson M.D. Computer Science and Engineering Computer Science and Engineering UW Medicine Sleep Center University of Washington University of Washington University of Washington 摘要：

2、我们在智能手机上提出一种检测睡眠呼吸暂停时间的非接触解决方法。为了做到这一点，我们引入一种智能手机监控呼吸引起的每分钟胸部和腹部运动的新系统。我们的系统运行在手机远离主体环境下并且能够同时确认和跟踪来自多个主体的细粒度呼吸运动。我们将手机变成一个主动声纳系统发射频率调制声音信号并且监听他们的反射。我们的设计监控这些反射的每分钟变化来提取胸部运动。家居环境的结果显示我们的设计有效地工作在达距离达1米甚至在毯子下的主体。建立上述的系统，我们开发出识别来自声纳反射中各种睡眠呼吸暂停事件的算法，包括阻塞性睡眠呼吸暂停、中枢神经性睡眠呼吸暂停、呼吸不足。我们在威斯康星大学医学院睡眠中心港景中部署我们的系

3、统并且执行一个有37位患者共296小时的临床实验。我们的研究证明我们的系统识别呼吸事件的数目与真实数据之间是高相关的，中枢神经性睡眠呼吸暂停、阻塞性睡眠呼吸暂停和呼吸不足的相关系数分别是0.9957、0.9860和0.9533。此外，在计算呼吸暂停和呼吸不足事件发生率的平均错误低于每小时1.9件。分类和主题描述J.3【计算机应用】：生命科学与医学概括词语设计；人的因素；算法关键词移动健康；睡眠呼吸暂停；手机声纳；非接触呼吸监控1. 介绍睡眠呼吸暂停是一种普通医学症状，在睡眠呼吸被打断时发生。估计影响超过1800万美国成年人并且被连接到关注缺陷/多障碍、高血压、糖尿病、心脏病、中风，它增加了机动

4、车辆事故。在临床中诊断睡眠呼吸暂停需要多导睡眠图测试，这是一个昂贵、耗时和劳动密集的过程。它需要训练有素的技术人员在睡眠期间中对患者连接和监控各种传感器，这是一个典型相关联的长时间的等待表。当便携式记录系统发展使用在家庭环境时，他们要求给病人或者病床安装大量的传感器并且大多数要求一个训练有素的技术人员设置记录系统。图1在多导睡眠图测试中使用传感器。该图显示在测试和一部分收集数据中使用传感器。多导睡眠图被用来诊断多种睡眠病症包括睡眠呼吸暂停。我们的目标是在身体上使用智能手机检测睡眠呼吸暂停不需要任何传感器。在这篇论文中我们提出一下的问题：我们能否不需要安装传感器利用智能手机检测睡眠呼吸暂停事件?

5、最主要的困难是检测睡眠呼吸暂停事件需要跟踪因为呼吸引起的细粒度腹部和胸部运动。虽然IPhone呼吸应用可以跟踪呼吸运动，它需要将手机放在胸腔和腹部之间，因此它是侵入是的。基于视觉解决方法能够不要安装用户就能跟踪这些运动，但是这仅限于视线范围和良好的光照条件，因此并不适用于睡眠环境，例如：在夜间或者在毯子下。我们引入一种新的在智能手机上跟踪胸部和腹部运动的非接触系统并且能在睡眠环境中工作。它能在手机远离使用者是工作并且同时跟踪来自多个使用者的呼吸运动。使用这个设计。我们建立AnpeaApp，一个基于智能手机在临床在多导睡眠图测试记录解决睡眠相关呼吸事件的解决方法，包括呼吸不足（当主体呼吸变浅）、

6、阻塞性睡眠呼吸暂停（主体上气道整个或部分阻塞）、中枢神经性睡眠呼吸暂停（当主体屏住呼吸）。我们关键的观点是将手机变成主动的声呐系统。在一个高水平下，我们从手机扬声器中发送18-20kHZ的声波并且在麦克风中监听他们的反射。胸部和腹部运动是因为呼吸产生的变化反射声波。这些变化是每分钟的，但是从其他环境中可靠地提取他们是困难的。为了克服这一点，我们在载波频率中采用FMCW(频率调制连续波)传输和移动，并且允许我们通过映像时间差将到达不同时间的反射分离。具体地说，来自人体的反射到达一个具体时间取决于到手机扬声器的距离。因此，在所有其他环境反射存在下着眼于相应的频率允许我们有效地提取因为呼吸引起的振幅

7、变化。此外，由于来自多个主体的反射在不同的时间内到达，相关频率为我们提供了同时跟踪多个呼吸信号的能力。最后，非呼吸身体运动产生的反射模式不同于呼吸产生的，能够让我们区分它们。我们基于现有的智能手机实现我们的设计并且执行在卧室环境下5个健康参与者使用游标呼吸带作为基线的基准实验。我们的结果如下显示：l 我们系统评估粗粒度呼吸频率是距离离主体达到1米时基线的99.2%之内。这种转换的误差小于0.11次/min。当主体裹着毯子时这些的准确度仍然很高。l 在环境中上述的准确度仍然不受来自附近街道上的车辆以及人们的交谈的可听噪声的影响。这是因为我们使用高通滤波器过滤出低于18kHZ的可听噪声。l 他能够

8、分开和同时跟踪在床上两个主体分开距离为20cm的呼吸运动。构建上述的系统，我们设计算法去计算中枢神经性睡眠呼吸暂停、阻塞性睡眠呼吸暂停和呼吸不足事件的数目以及呼吸暂停呼吸不足指数，它是一个在睡眠期间呼吸暂停和呼吸不足事件的平均速率。我们通过因为胸部和腹部运动以及非呼吸身体运动引起细粗粒度变化的过程获得这种指数。我们在威斯康星大学医学院睡眠中心港景部署ApneaApp并且执行一个37个病患共296小时的临床实验。在我们的研究中这些患者被他们的医师要求做多导睡眠图（PSG）测试。我们的研究与PSG测试同时进行并且我们认为和从后者的诊断作为真实数据来评估我们的系统。我们的研究如下：l 所有患者通过我

9、们系统的检测出中枢神经性睡眠呼吸暂停、呼吸不足和阻塞性睡眠呼吸暂停事件的数目与真实数据是高相关的。具体地，中枢神经性睡眠呼吸暂停、呼吸不足和阻塞性睡眠呼吸暂停在PSG和ApneaApp之间组内相关系数分别是0.9957、0.9533和0.9860。l 在计算呼吸暂停和呼吸不足事件速率的平均误差是1.9件/hr；这是一个临床可接受的值。l 我们的系统准确地从37位病患中将32个分类为四个睡眠呼吸暂停等级（无呼吸暂停、轻度、中度和深度呼吸暂停）。5个错误分类发生在无呼吸暂停和轻度呼吸暂停之间；他们中的4个正好发生在两个等级的边界并且误差小于1件/hr。这些边界事件被医生通过病患的偏好、症状和保险分

10、别处理；因此能够有效地将错误分类的数目减少到1个。l 我们执行一个87位睡眠呼吸暂停患者和在威斯康星大学CSE的57名本科学生的可听测试，在ApneaApp中87名睡眠呼吸暂停患者没有一个报告任何可听声音。57名本科生中有两名报告听到可听声音。这个表明ApneaApp对大多数成年人是不可听的。贡献：我们做出四个贡献：(1)我们引入一种新兴的在智能手机上跟踪因为呼吸引起的胸部和腹部运动的非接触技术。我们通过用FMCW声纳传输分析反射来获得。(2)我们设计算法检测中枢神经性睡眠呼吸暂停、阻塞性睡眠呼吸暂停和呼吸不足以及评估来自声纳反射的呼吸暂停呼吸不足指数。(3)我们在现有的智能手机实现我们的设计

11、并且证明能够同时跟踪来自多个主体的呼吸运动的能力。(4)我们执行一个37名病患的临床研究证明我们系统准确计算中枢神经性睡眠呼吸暂停、阻塞性睡眠呼吸暂停和呼吸不足的数目以及呼吸暂停呼吸不足指数的有效性。Respiratory Summary:Types of Respiratory Events/Respiratory Effort Related Arousal(RERA) EventsRespiratory EventsNumberIndexRERAParameterTotalIndexObstructive Anpeas12315.7/hrTotal:12916.4Mixed Apneas

12、00.0/hrNon-REM:10516.5Central Apneas10.1/hrREM:2416.1Total Apneas12415.8/hrSupine:12916.4Total Hypopneas*37147.2/hrLateral:N/AN/AApnea + Hypopneas*49563.0/hrProne:N/AN/A图2一个临床PSG报告的快照。它总结了随着呼吸暂停呼吸不足指数(AHI)阻塞性睡眠呼吸暂停、中枢神经性睡眠呼吸暂停和呼吸不足事件的数目。一个AHI值在0-5之间被分成无呼吸暂停，值在5-15之间被分成轻度呼吸暂停，AHI值在15-30之间被分成中度呼吸暂停，和更

13、高的AHI值是深度呼吸暂停状态。2. 多导睡眠图概述临床多导睡眠图测试是传统用来诊断睡眠呼吸暂停和其他睡眠病症。它是在睡眠实验室进行整夜，那里有训练有素的技术人员监控患者睡眠模式。为了做到这一点，技术人员为患者加上许多传感器包括一个测试呼吸运动的胸部和腹部带、一个鼻压传感器、一个打鼾麦克风、一个测量血氧饱和度的脉搏血氧仪、在每个腿上一个运动传感器检测运动和5个脑电传感器测量脑部运动。这些传感器都是用电线连接并且在整个睡眠期间技术人员监控来自于传感器的生命数据流。图2显示一个PSG报告的快照。主要的度量用来诊断呼吸暂停的是AHI 呼吸暂停呼吸不足指数它表示在睡眠期间呼吸暂停和呼吸不足发生的速率。

14、医师用AHI的值将睡眠呼吸暂停划分等级。具体地，AHI值在0-5之间被划分成无呼吸暂停，在5-15之间划分成轻度呼吸暂停，AHI值在15-30之间划分成中度呼吸暂停和更高的AHI值是深度呼吸暂停。呼吸暂停呼吸不足指数计算如下：AHI=#central apnea+#hypopnea+#obstructive apneatotal sleep time在上述等式中，central apnea、hypopnea和obstructive apnea表示在研究期间被跟踪的各种呼吸暂停状态。混合呼吸暂停是呼吸暂停的另一种，并且在上述等式中有时包括在内的。然而，在我们的PSG中显示非零的混合呼吸暂停是没有

15、的，所以在我们的计算中会忽略他们。为了计算上述的等式，8小时传感器数据被分成30s的间隔称为时期。分析这些时期的评分过程涉及两步。第一步是状态，在每个时期中确认病人是醒着还是睡着。这个可以通过5个脑电传感器得到测试脑部活动来获得。在这一步的结尾，每个时期被标注着醒着或者睡着的状态。第二步包含确认中枢神经性睡眠呼吸暂停、呼吸不足和阻塞性睡眠呼吸暂停事件的数目，以下是使用AASM准则的概要。图3美国睡眠医学学院（AASM）睡眠呼吸暂停信号描述。该图显示三种睡眠呼吸暂停的胸部运动和鼻腔压力信号。当主体屏住呼吸一段时间时中枢神经性睡眠呼吸暂停事件发生。当主体胸部运动下降超过30%同时有3-4%的血氧饱

16、和度时呼吸不足事件发生。最后，当主体努力将空气送到肺部但因为阻塞空气不能到达肺部时阻塞性睡眠呼吸暂停事件发生。确认中枢神经性睡眠呼吸暂停事件。当主体屏住呼吸一段时间时中枢神经行睡眠呼吸暂停事件发生。图3所示在中枢神经行睡眠呼吸暂停事件中鼻腔压力和胸部运动信号。该图显示在努力呼吸中胸部运动不存在平坦直线；必然地鼻腔压力也是平坦的。如果这个持续超过10s，这段被标记成中枢神经行睡眠呼吸暂停事件。确认呼吸不足事件。当主体呼吸变浅时呼吸不足事件发生。图3显示在呼吸不足事件期间鼻腔压力和胸部运动信号。该图显示在呼吸不足期间，胸部运动减少振幅。尤其是，如果振幅减少超过30%并且同时又3-4%的血氧饱和度，

17、然后相应的期间被标记成呼吸不足。我们注意到现有的临床研究显示30%的减少可单独检测呼吸不足没有明显的精度减少。确认阻塞性睡眠呼吸暂停事件。在睡眠期间上气管整个或者部分堵塞。图3显示胸部带信号中可看见的努力呼吸的信号以及在鼻腔压力传感器中空气流是平缓的。我们注意下面关于PSG的三点：l 当前程序是传感器数据收集以及劳动和时间紧密的过程。具体地，技术人员要花费一个小时给每位患者安装许多传感器。在整个8小时睡眠期间，技术人员监控这些传感器并且确保它们正确的连接在患者的身体上。传感器数据在手工处理后标记睡眠呼吸暂停事件的每个时期。l 尽管便携式睡眠呼吸暂停测试在家里执行，它仍然需要设置胸部和腹部带、鼻

18、腔压力传感器、热敏电阻传感器、EKG和脉搏血氧仪。家庭测试有高失败率达到33%因为电线脱落导致信号缺失。l 一个PSG测试也可以用来诊断其他睡眠相关状态包括上气道阻力综合症，它包含努力呼吸相关觉醒(RERA)，在图2. RERAs是不符合上述睡眠觉醒呼吸暂停和呼吸不足的定义。尽管这些是与呼吸相关的并且可以用来检测我们基于声纳的系统，详细地检测他们不在本文的范围之内。2.外文原文Contactless Sleep Apnea Detection on SmartphonesRajalakshmi Nandakumar Shyamnath Gollakota Nathaniel Watson M.

19、D. Computer Science and Engineering Computer Science and Engineering UW Medicine Sleep Center University of Washington University of Washington University of Washington Abstract We present a contactless solution for detecting sleep apnea events on smartphones.

20、To achieve this, we introduce a novel system that monitors the minute chest and abdomen movements caused by breathing on smartphones. Our system works with the phone away from the subject and can simultaneously identify and track the fine-grained breathing movements from multiple subjects. We do thi

21、s by transforming the phone into an active sonar system that emits frequency-modulated sound signals and listens to their reflections; our design monitors the minute changes to these reflections to extract the chest movements. Results from a home bedroom environment shows that our design operates ef

22、ficiently at distances of up to a meter and works even with the subject under a blanket. Building on the above system, we develop algorithms that identify various sleep apnea events including obstructive apnea, central apnea, and hypopnea from the sonar reflections. We deploy our system at the UW Me

23、dicine Sleep Center at Harborview and perform a clinical study with 37 patients for a total of 296 hours. Our study demonstrates that the number of respiratory events identified by our system is highly correlated with the ground truth and has a correlation coefficient of 0.9957, 0.9860, and 0.9533 f

24、or central apnea, obstructive apnea and hypopnea respectively. Furthermore, the average error in computing of rate of apnea and hypopnea events is as low as 1.9 events/hr.CATEGORIES AND SUBJECT DESCRIPTORSJ.3 Computer Applications: Life and Medical SciencesGENERAL TERMSDesign; Human Factors; Algorit

25、hmsKEYWORDSMobile Health; Sleep Apnea; Phone Sonar; Contactless Breathing Monitoring1. INTRODUCTIONFigure 1Sensors used in the polysomnography test. The figure shows all the sensors used in the test along with the data collection unit. Polysomnography is used to diagnose various sleep disorders incl

26、uding sleep apnea. Our goal is to use a smartphone to detect sleep apnea events without any sensors on the human body.Sleep apnea is a common medical disorder that occurs when breathing is disrupted during sleep. It is estimated to affect more than 18 million American adults and is linked to attenti

27、on deficit/hyperactivity disorder, high blood pressure, diabetes, heart attack, stroke, and increased motor vehicle accidents. Diagnosing sleep apnea in the clinic requires the polysomnography test which is an expensive, time-consuming and labor-intensive process. It requires a trained technician to

28、 attach and monitor various sensors on the patient for the sleep duration and is typically associated with long waiting lists. While portable recording systems are being developed for use in home settings, they require instrumenting either the patientor the bed with various sensors and most still re

29、quire a trained technician to setup the recording system .In this paper we ask the following question: Can we leverage smartphones to detect sleep apnea events without the need for sensor instrumentation? The key challenge is that detecting sleep apnea events requires tracking the fine-grained abdom

30、en and chest movements due to breathing. While the iPhone Respiratory app can track the breathing movements, it requires placing the phone on the body between the ribcage and the stomach and hence is intrusive. Vision-based solutions can track these movements without instrumenting users, but are lim

31、ited to line-of-sight and good lighting conditions and hence are not applicable to the sleep environment, i.e., in the dark or under a blanket. We introduce a novel contactless system that tracks the chest and abdomen movements on smartphones and works in the sleep environment. It operates with the

32、phone away from the user and can concurrently track the breathing movements from multiple users. Using this design, we build ApneaApp, a smartphone-based solution for detecting sleep-related respiratory events reported in a clinical polysomnography test including hypopnea (when the subjects breathin

33、g becomes shallow), obstructive apnea (a complete or partial blockage of the subjects airway) and central apnea (when the subject holds his or her breath). Our key insight is to transform the phone into an active sonar system. At a high level, we transmit 18-20 kHz sound waves from the phone speaker

34、 and listen to their reflections at the microphone. The chest and abdomen motion due to breathing creates changes to the reflected sound waves. These changes, however, are minute and extracting them reliably from other environmental reflections is challenging. To overcome this, we employ FMCW (frequ

35、ency modulated continuous wave) transmissions that allow us to separate reflections arriving at different times by mapping time differences to shifts in the carrier frequency. Specifically, the reflections from the human body arrive at a specific time depending on the distance from the phone speaker

36、. Thus, focusing on the corresponding frequency allows us to reliably extract the amplitude changes due to breathing, in the presence of all other environmental reflections. Further, since reflections from multiple subjects would arrive at different times, the corresponding frequencies provide us wi

37、th the ability to simultaneously track multiple breathing signals. Finally, non-breathing body motion creates reflection patterns distinct from breathing, enabling us to distinguish between them. We implement our design on off-the-shelf smartphones and run benchmark experiments with five healthy par

38、ticipants in a bedroomenvironment using the Vernier respiratory belt as a baseline. Ourresults show the following: Our system estimates the coarse-grained breathing frequency1 to within 99.2% of the baseline at distances of up to a meter from the subject. This translated to an error of less than 0.1

39、1 breaths/min. These accuracies remain this high even when the subjects use blankets. The above accuracies remain unaffected by audible noise in the environment from vehicles on a nearby street as well as human conversations. This is because, we use a high-pass filter to filter out audible signals b

40、elow 18 kHz. It can separate and concurrently track the breathing movements of two subjects on the bed separated by 20 cm.Building on the above system, we design algorithms to compute the number of central, obstructive and hypopnea events as well as the apnea-hypopnea index which is the average rate

41、 of apnea and hypopnea events during the sleep duration. We achieve this by processing both the fine- and coarse- grained changes due to the chest and abdomen movements as well as non-breathing body motion. We deploy ApneaApp at the UW Medicine Sleep Center at Harborview and perform a clinical study

42、 with 37 patients for a total of 296 hours. The patients in our study were ordered by their physicians to undergo the polysomnography (PSG) test. Our study was done concurrently with the PSG test and we consider the sensor data and diagnosis from the latter as the ground truth for evaluating our sys

43、tem. Our study shows the following: Across patients, the number of central apnea, hypopnea and obstructive apnea events detected by our system is highly correlated with the ground truth. Specifically, the intra-class correlation coefficient between PSG and ApneaApp, is 0.9957, 0.9533 and 0.9860 for

44、central apnea, hypopnea and obstructive apnea respectively. The average error in computing the rate of apnea and hypopnea events is 1.9 events/hr; this is a clinically acceptable value . Our system accurately classifies 32 out of 37 patients between four sleep apnea levels (no apnea, mild, moderate,

45、 and severe apnea). The five misclassifications occur between no-apnea and mild-apnea; four of them happen right at the boundary between the two levels with an error less than 1 event/hr. These boundary cases are handled separately by physicians depending on the patient preferences, symptoms, and in

46、surance; thus, effectively reducing the number of misclassifications to one. We ran an audibility test with 87 sleep apnea patients (ages between 23 and 93 with a mean age of 50) and 57 healthy undergraduate students at UW CSE. None of the 87 sleep apnea patients reported any audible sounds from Apn

47、eaApp. Only two of the 57 undergraduates reported hearing audible sounds. This demonstrates that ApneaApp is inaudible for most of the adult population.Figure 2Snapshot of a Clinical PSG Report. It summarizes the number of obstructive, central and hypopnea events along with the apneas-hypopneas inde

48、x (AHI). An AHI value between 05 is classified as no-apnea, values between 515 are classified as mildapnea, AHI values between 1530 are classified as moderate-apnea, and higher AHIs are severe apnea conditions.Contributions: We make four key contributions: (1) We introduce a novel contactless techni

49、que for tracking chest and abdomen movements due to breathing on smartphones. We achieve this by analyzing the reflections from FMCW sonar transmissions. (2) We design algorithms to detect central apnea, obstructive apnea, and hypopnea as well as estimate the apnea-hypopnea index from the sonar refl

50、ections. (3) We implement our design on off-the-shelf smartphones and demonstrate the ability to concurrently track breathing movements from multiple subjects. (4) We perform a clinical study with 37 patients demonstrating the feasibility of using our system to accurately compute the number of centr

51、al, obstructive, and hypopnea events as well as the apnea-hypopnea index.2. POLYSOMNOGRAPHY OVERVIEWThe clinical polysomnography test (PSG) is traditionally used to diagnose sleep apnea and other sleep disorders. It is conducted overnight in a sleep laboratory where a trained technician monitors the

52、 patients sleeping patterns. To do this, the technician attaches the patient with a number of sensors including a chest and abdomen belt to measure breathing movements, a nasal pressure transducer, a snore microphone, a pulse oximeter to measure oxygen saturation, a movement sensor on each leg to de

53、tect movements and five EEG sensors to measure brain activity. The sensors are all connected using wires and the technician monitors the live data stream from the sensors, throughout the sleep duration. Fig. 2 shows a snapshot of a PSG report. The key metric used for sleep apnea diagnosis is the AHI

54、 the Apnea-Hypopnea Index that represents the rate at which apnea and hypopnea events occur during the sleep period. Physicians classify the sleep apnea level using these AHI values. Specifically, AHI values between 05 are classified as no-apnea, those between 515 are classified as mild-apnea, AHI v

55、alues between 1530 are classified as moderate-apnea, and higher AHIs are severe apnea. Figure 3American Academy of Sleep Medicine (AASM) Signal Characterization of the Apnea Events. The figures show the chest motion and nasal pressure signals for the three apneas. A central apnea event occurs when t

56、he subject holds her breath for a nonnegligible duration. A hypopnea event occurs when the subjects chest motion drops by more than 30% with an accompanying 3 4% oxygen desaturation. Finally, an obstructive apnea event occurs when the subject makes an increased effort to pull air into the lungs but

57、air does not reach the lungs due to blockage.The apnea-hypopnea index is computed as follows:In the above equation, central apnea, hypopnea, and obstructive apnea denote the various apnea conditions that are tracked during the study. Mixed apneas are another class of apneas that are sometimes includ

58、ed in the above equation. However, none of our PSG reports showed non-zero mixed apneas and so we ignore them in our computation. To compute the above parameters, the eight-hour sensor data is split into 30-second intervals called epochs. The scoring process of analyzing these epochs involves two main steps. The first step is staging, which identifies whether the patient is awake or asleep in each epoch. This is achieved by examining the brain activity obtained from the five EEG sensors. At the end of this step, each epoch is marked as being in either a wake or sleep stage. The second ste