本文選題：心電信號(hào)R波　切入點(diǎn)：HRV信號(hào)　出處：《中國人民解放軍軍醫(yī)進(jìn)修學(xué)院》2007年碩士論文

【摘要】： 心電信號(hào)(ECG)是心臟電活動(dòng)在體表的綜合表現(xiàn)。心電信號(hào)處理的目的是根據(jù)心電信號(hào)的特征推斷心血管系統(tǒng)的狀態(tài)，并據(jù)之做出輔助的醫(yī)學(xué)決策。本論文以心電信號(hào)為研究對(duì)象，主要包含以下三個(gè)方面的內(nèi)容： (1)在研究了現(xiàn)有的一些ECG信號(hào)中的QRS波群檢測方法后，提出了一種算法簡單、速度快的適合于實(shí)時(shí)信號(hào)檢測的基于一、二階差分相結(jié)合的改進(jìn)的差分閾值算法。該算法的原理是通過一、二階差分運(yùn)算并結(jié)合取平方的非線性運(yùn)算，最大限度地抑制各類噪聲同時(shí)提高QRS波群的份量。運(yùn)用本算法對(duì)原始心電數(shù)據(jù)文件、加入噪聲的心電數(shù)據(jù)和MIT心電數(shù)據(jù)庫的數(shù)據(jù)進(jìn)行分析，可以極大地抑制干擾而準(zhǔn)確地檢測R波波峰位置，，驗(yàn)證了本算法的有效性。 (2)分析由原始ECG得到的R-R間期即HRV信號(hào)，是一種非侵入式的研究自主神經(jīng)系統(tǒng)功能的途徑。經(jīng)過對(duì)多種HRV信號(hào)的分析方法的研究，本文選擇了有代表性的并適用于短程(約5min)數(shù)據(jù)的分析方法和計(jì)算參數(shù)：a、基于時(shí)域的平均心率(AHR)和相鄰R-R間期差值的均方根值(r-MSSD)b、基于頻域的總功率(TP)、歸一化低頻、高頻功率(LFnorm、HFnorm)、低頻和高頻的功率比(LF／HF) c、基于非線性動(dòng)力學(xué)的分形維數(shù)(FD)、近似熵(ApEn)和建模非線性方程的NAR模型方程的總項(xiàng)數(shù)(TL)對(duì)從RS-621多導(dǎo)睡眠儀獲得的處于兩個(gè)特定的睡眠階段：NREM4期(非快速眼動(dòng)4期)和REM期(快速眼動(dòng)期)的10個(gè)健康成年男子的HRV數(shù)據(jù)進(jìn)行分析，得到的實(shí)驗(yàn)結(jié)論是，NREM4期的HRV數(shù)據(jù)的各項(xiàng)指標(biāo)均顯著低于REM期的對(duì)應(yīng)值。推斷產(chǎn)生這樣的實(shí)驗(yàn)結(jié)果是由于REM期較NREM4期，心血管系統(tǒng)的復(fù)雜性更高、調(diào)節(jié)機(jī)制更加復(fù)雜，這個(gè)實(shí)驗(yàn)結(jié)論和以往生理實(shí)驗(yàn)得到的結(jié)論具有一致性。 (3)信號(hào)的奇異性往往攜帶著信號(hào)的重要特征信息。通過證明小波變換模極大值和信號(hào)的奇異性指標(biāo)-Lipschitz指數(shù)的關(guān)系定理|T_ψ(a，b)|_(b=t_0)≤Ca~(a(t_0)+1/2)后，提出并實(shí)現(xiàn)了在MATLAB中求解Lipschitz指數(shù)的算法，并將之用于正常人和來自MIT心電數(shù)據(jù)庫中的心律不齊病人的ECG信號(hào)的R波計(jì)算中，得到了正常人R波的奇異性顯著高于心律失常病人的結(jié)論，推斷心律失常病人奇異性的降低可能是處在病理狀態(tài)的心臟系統(tǒng)的調(diào)節(jié)能力下降，復(fù)雜性降低導(dǎo)致的。通過計(jì)算R波的Lipschitz指數(shù)，從數(shù)理的角度對(duì)不同模式R波波形的奇異性給出了量化指標(biāo)，為心電信號(hào)分析的計(jì)算機(jī)化提供了一個(gè)思路。 本文主要以R波檢測、基于R-R間期的HRV特征信號(hào)分析和R波的Lipschitz指數(shù)的計(jì)算作為文章的論述重點(diǎn)，此外還做了一些基于LabVIEW環(huán)境實(shí)現(xiàn)的心電數(shù)據(jù)采集方面的工作，實(shí)現(xiàn)了基于普通電腦聲卡和基于USB6009數(shù)據(jù)采集卡的簡單心電采集系統(tǒng)。通過對(duì)比發(fā)現(xiàn)，基于聲卡的采集系統(tǒng)對(duì)數(shù)據(jù)源有一定的匹配要求，信號(hào)較容易產(chǎn)生失真，并且采樣率是四個(gè)固定的默認(rèn)值，因此系統(tǒng)的實(shí)際應(yīng)用價(jià)值不大；而基于USB6009的系統(tǒng)可以實(shí)現(xiàn)多通道多路信號(hào)的實(shí)時(shí)高速采集、顯示和保存，具有很高的普適性，適于構(gòu)建方便的數(shù)據(jù)采集平臺(tái)，為實(shí)驗(yàn)室環(huán)境下的信號(hào)研究和分析提供了很好的數(shù)據(jù)來源的途徑。
[Abstract]:Cardiac electrical signal ( ECG ) is an integrated representation of cardiac electrical activity in the body surface . The purpose of ECG signal processing is to infer the state of cardiovascular system based on the characteristics of ECG signal and to make auxiliary medical decision .









( 1 ) After studying the QRS complex detection method in some existing ECG signals , an improved differential threshold algorithm is proposed which is suitable for real - time signal detection , which is suitable for real - time signal detection .









( 2 ) Analysis of HRV data of 10 healthy adult men derived from the original ECG , i.e . HRV signal , is a non - invasive approach to the study of autonomic nervous system function . The results of this study are as follows : a . Mean Square Root Mean ( r - MSSD ) b , based on Time Domain Average Heart Rate ( AHR ) and Adjacent R - R Interval Difference ( TL ) . Based on the frequency domain total power ( TP ) , normalized low frequency , high frequency power ( LFnorm , HFnorm ) , approximate entropy ( ApEn ) , and model nonlinear equation , the total number of items ( TL ) is higher than that of REM period . It is concluded that the results of this experiment are consistent with the conclusion of the previous physiological experiment .









( 3 ) The singularity of the signal often carries important characteristic information of the signal . After proving the relationship of the maximum value of wavelet transform modulus and the singular index of the signal , the relation theorem of the exponent of the singular value of the wavelet transform modulus and the singular index of the signal ( b = t _ 0 ) 鈮

本文編號(hào)：1700974