本文選題：睡眠分期 + 心率變異性��； 參考：《電子科技大學(xué)》2017年碩士論文

【摘要】：越來越多的人開始遭遇睡眠問題,評價(jià)睡眠質(zhì)量進(jìn)而改善睡眠狀況已經(jīng)成為一大課題。準(zhǔn)確的睡眠分期是客觀評估睡眠質(zhì)量和診斷睡眠相關(guān)疾病的基礎(chǔ),經(jīng)典自動(dòng)分期方法基本上是對腦電信號進(jìn)行分析的。但是,腦電信號的記錄操作復(fù)雜,成本高,電極的放置也會(huì)干擾人的正常睡眠,無法滿足家庭睡眠監(jiān)測場合的需要。人們提出用較為方便檢測的生理參數(shù)去自動(dòng)分析睡眠過程,相較傳統(tǒng)腦電更具實(shí)用價(jià)值,利用壓電感知床墊能實(shí)現(xiàn)人在睡眠時(shí)多個(gè)生理參數(shù)的長時(shí)間同步檢測,對睡眠幾乎沒有任何干擾,可以在家庭中監(jiān)測人們的真實(shí)睡眠狀況,擁有良好的應(yīng)用前景。床墊采集的是心沖擊圖(BCG)信號,包含心率、呼吸率和體動(dòng)信息,但目前對BCG信號進(jìn)行自動(dòng)睡眠分期的準(zhǔn)確率不高。本文的研究目的是基于BCG信號建立一種更加準(zhǔn)確可靠的多參數(shù)自動(dòng)分期算法,然后應(yīng)用于實(shí)現(xiàn)的床墊式實(shí)時(shí)睡眠監(jiān)護(hù)系統(tǒng)中,進(jìn)行睡眠的實(shí)時(shí)監(jiān)測。本文采用從BCG信號中計(jì)算出的心率、呼吸和體動(dòng)序列進(jìn)行自動(dòng)睡眠分期,分為四個(gè)階段,即覺醒期、淺度睡眠期、深度睡眠期和快速眼動(dòng)睡眠期。由于不同睡眠階段的心率波形形態(tài)較難區(qū)分,利用時(shí)變自回歸模型進(jìn)行心率變異性(HRV)的特征提取,進(jìn)一步基于隱馬爾可夫模型進(jìn)行特征的自動(dòng)分類識別。發(fā)現(xiàn)高頻段極點(diǎn)相位與總功率的特征組合能夠較好地區(qū)分各睡眠分期,并結(jié)合呼吸率和體動(dòng)信息校正分期結(jié)果,能夠有效地提高基于HRV的分期準(zhǔn)確率。使用MIT-BIH數(shù)據(jù)庫中的數(shù)據(jù)測試,比較文中算法和專家的分期結(jié)果,驗(yàn)證了建立的多參數(shù)分期算法的精度,識別率達(dá)到70.13%,且計(jì)算快速,可用于實(shí)時(shí)監(jiān)測。本文設(shè)計(jì)并實(shí)現(xiàn)了床墊式實(shí)時(shí)睡眠監(jiān)護(hù)系統(tǒng),將本文建立的自動(dòng)睡眠分期算法應(yīng)用到該系統(tǒng)中,能在家庭環(huán)境中使用,實(shí)時(shí)監(jiān)測人們的睡眠。系統(tǒng)基于壓電感知床墊,將從BCG信號分離出的實(shí)時(shí)的心率、呼吸率和體動(dòng)數(shù)據(jù)上傳至服務(wù)器進(jìn)行存儲(chǔ),完成自動(dòng)睡眠分期,最終通過智能手機(jī)應(yīng)用為用戶提供實(shí)時(shí)睡眠信息展示、睡眠質(zhì)量分析和改善建議等服務(wù),以幫助人們提高睡眠質(zhì)量。
[Abstract]:More and more people are experiencing sleep problems. It has become a major topic to evaluate sleep quality and improve sleep quality. Accurate sleep staging is the basis of objective evaluation of sleep quality and diagnosis of sleep related diseases. However, the recording of EEG signals is complicated and costly, and the placement of electrodes will interfere with normal sleep, which can not meet the needs of family sleep monitoring. More convenient physiological parameters are proposed to automatically analyze sleep process, which is more practical than traditional EEG. Using piezoelectric sensing mattress can realize long time synchronous detection of multiple physiological parameters during sleep. It can monitor people's real sleep condition in the family, and has good application prospect. The mattresses collected the BCG) signals of cardiogram, which included heart rate, respiration rate and body movement information, but the accuracy of automatic sleep staging of BCG signals was not high at present. The purpose of this paper is to establish a more accurate and reliable multi-parameter automatic staging algorithm based on BCG signal, and then apply it to the realization of the mattress real-time sleep monitoring system to monitor sleep in real time. In this paper, the heart rate, respiration and body motion sequences calculated from BCG signal are used to carry out the automatic sleep stages, which are divided into four stages, namely, wakefulness, shallow sleep, deep sleep and rapid eye movement sleep. Because it is difficult to distinguish the shape of heart rate waveform in different sleep stages, the time-varying autoregressive model is used to extract the feature of HRV, and then the hidden Markov model is used for automatic classification and recognition. It is found that the feature combination of pole phase and total power in high frequency band can distinguish the sleep stages well, and can effectively improve the accuracy of staging based on HRV by combining the results of respiration rate and volume motility information correction. By using the data test in MIT-BIH database and comparing the results of the algorithm and the expert, the accuracy of the multi-parameter staging algorithm is verified. The recognition rate is 70.133.And the calculation is fast and can be used for real-time monitoring. In this paper, a mattress type real-time sleep monitoring system is designed and implemented. The automatic sleep staging algorithm established in this paper is applied to the system, which can be used in the home environment and monitor people's sleep in real time. Based on piezoelectric sensing mattress, the system uploads the real-time heart rate, respiration rate and body motion data separated from BCG signal to the server for storage, and completes the automatic sleep stage. Finally, the smartphone application provides users with real-time sleep information display, sleep quality analysis and improvement advice to help people improve the quality of sleep.
【學(xué)位授予單位】：電子科技大學(xué)
【學(xué)位級別】：碩士
【學(xué)位授予年份】：2017
【分類號】：TP311.52;R740

【參考文獻(xiàn)】

相關(guān)期刊論文 前10條

1 李延軍;仲崇發(fā);李琳;祝瑞云;;一種僅使用呼吸信號檢測非眼動(dòng)睡眠的方法[J];航天醫(yī)學(xué)與醫(yī)學(xué)工程;2015年04期

2 王德璽;張宗平;劉紅;李韻;雷飛;唐向東;;體動(dòng)記錄儀在睡眠和睡眠障礙監(jiān)測中的應(yīng)用[J];生物醫(yī)學(xué)工程學(xué)雜志;2014年01期

3 曹欣榮;王昆;張晶;唐勁天;;心沖擊圖心率變異性分析的可行性[J];科技導(dǎo)報(bào);2014年Z1期

4 蔣芳芳;王旭;楊丹;;心沖擊圖信號中呼吸成分的時(shí)頻檢測方法研究[J];生物醫(yī)學(xué)工程學(xué)雜志;2012年03期

5 張莉莉;葛宏;呂曉東;胡曉林;蔣科;范軍;;基于體動(dòng)信號的睡眠、覺醒識別方法研究[J];醫(yī)療衛(wèi)生裝備;2011年04期

6 王海濤;鄭慧君;曹征濤;楊軍;俞夢孫;;考慮個(gè)體特征的非腦電睡眠分期[J];中國生物醫(yī)學(xué)工程學(xué)報(bào);2010年02期

7 王菡僑;;有關(guān)美國睡眠醫(yī)學(xué)學(xué)會(huì)睡眠分期的最新判讀標(biāo)準(zhǔn)指南解析[J];診斷學(xué)理論與實(shí)踐;2009年06期

8 熊國良;張龍;陳慧;;TVAR在非平穩(wěn)工況轉(zhuǎn)子故障診斷中的應(yīng)用[J];振動(dòng)、測試與診斷;2007年02期

9 寧艷;江朝暉;安濱;馮煥清;;睡眠生理參數(shù)的去趨勢波動(dòng)分析[J];生物醫(yī)學(xué)工程學(xué)雜志;2007年02期

10 莊志;高上凱;高小榕;;基于心率變異分析的睡眠分期方法[J];生物醫(yī)學(xué)工程學(xué)雜志;2006年03期

相關(guān)碩士學(xué)位論文 前5條

1 程佳;基于腦電信號的睡眠分期研究[D];北京理工大學(xué);2015年

2 馮曉明;基于腕動(dòng)信號的睡眠質(zhì)量監(jiān)測裝置設(shè)計(jì)[D];華南理工大學(xué);2014年

3 張晶;壓電薄膜式心沖擊圖信號采集系統(tǒng)的開發(fā)與應(yīng)用[D];河北工業(yè)大學(xué);2014年

4 崔娜;微動(dòng)敏感床墊式睡眠監(jiān)測儀在OSAHS診斷中的應(yīng)用[D];吉林大學(xué);2011年

5 梁曉花;基于腦電心電數(shù)據(jù)融合的睡眠分期[D];江蘇大學(xué);2008年

，

本文編號：1840012