基于动态心电质量分析的R波检测研究与实现

发布时间：2018-04-01 23:31

本文选题：动态心电　切入点：运动伪迹　出处：《东北大学》2012年硕士论文

【摘要】：近年来,心电信号的处理和应用重点逐渐转向日常生活中行为下的检测,而R波提取方法的研究仍一直被作为心电监测的重点。然而,在长时间和动态的心电信号记录过程中,将引入很多噪声。其中最为常见的噪声即肌电干扰,它严重影响R波的提取质量。本文通过采集动态心电信号,分析含有严重肌电干扰的动态心电信号的特征。分析表明,不同种类的肌肉运动会产生不同形态的肌电,在动态心电信号中一般存在两种肌电干扰信号：一种是肌肉细胞收缩或拉伸时产生的电信号,这种信号的特点是在幅值上快速的变化,能量基本分布在低频区域；另一种是肌肉细胞保持不动持续发力时产生的电信号,其特点是比前一种肌电信号频率高,幅值小,波动性较强,对心电信号中的R波有很大的干扰作用。本研究首先采集不同肌电干扰影响下的动态心电波形,然后通过频毒能量比、短时过阈值率和短时能量三个参数进行质量分析。其中,频率能量比参数可以用来区分静态心电信号、含持续性肌电干扰的心电信号和含突发性肌电干扰的心电信号,但是,它需要的帧长较长,对质量分析结果的准确性有一定的影响。短时过阈值率参数可以区分静态心电信号和动态心电信号,短时能量参数可以区分含持续性肌电干扰和含突发性肌电干扰的心电信号,它们的帧长较短,可以弥补频率能量比的帧长缺陷。结合时域和频域三种特征参数的结果,可实现短时内的动态心电信号质量分析。 R波检测方法的选择与质量分析结果有关。对含持续性肌电干扰的动态心电信号采用Dmey小波模极大值的多尺度分析方法。动态心电信号的变化趋势可以由经验模态分解法获得,所以对含突发性肌电干扰的动态心电信号采用本征模函数的多尺度分析进行R波检测。本文提出的基于质量分析的动态心电信号中R波检测方法由实验室中采集得到的数据样本进行验证,分析结果显示本方法在动态心电信号的分析与R波提取结果具有可接受的和可靠的准确性,基本可达到94%以上。
[Abstract]:In recent years, the emphasis of ECG signal processing and application has gradually shifted to the detection of the behavior in daily life, while the research of R wave extraction method has been regarded as the focus of ECG monitoring.However, a lot of noise will be introduced in the long and dynamic ECG recording process.The most common noise is myoelectric interference, which seriously affects the quality of R wave extraction.In this paper, the characteristics of dynamic ECG signals with serious EMG interference are analyzed by collecting dynamic ECG signals.Analysis shows that different kinds of muscle movements produce different forms of electromyography. There are generally two kinds of EMG interference signals in dynamic ECG signals: one is electrical signals produced by muscle cells when they contract or stretch.This signal is characterized by a rapid change in amplitude and the distribution of energy in the low frequency region; the other is the electrical signal produced when the muscle cell keeps its motionless and sustained force, which is characterized by a higher frequency and smaller amplitude than the previous EMG signal.It has strong volatility and has great interference to R wave in ECG signal.In this study, the dynamic ECG waveforms under the influence of different EMG interference were collected firstly, and then the quality analysis was carried out by three parameters: the ratio of frequency to poison energy, the short time threshold rate and the short time energy.Among them, the frequency-energy ratio parameters can be used to distinguish between static ECG signals, ECG signals with persistent EMG interference and ECG signals with sudden EMG interference, but it requires a longer frame length.It has a certain effect on the accuracy of the quality analysis results.It can make up for the frame length defect of the frequency energy ratio.Combined with the results of three characteristic parameters in time domain and frequency domain, the quality analysis of dynamic ECG signals in short time can be realized.The selection of R wave detection method is related to the result of quality analysis.The multiscale analysis method of Dmey wavelet modulus maximum is used for the dynamic ECG signal with persistent EMG interference.The change trend of dynamic ECG signal can be obtained by empirical mode decomposition method, so R wave detection of dynamic ECG signal with burst EMG interference is carried out by means of multi-scale analysis of eigenmode function.The method of R wave detection in dynamic ECG signal based on quality analysis proposed in this paper is verified by the data samples collected in the laboratory.The analysis results show that the method has acceptable and reliable accuracy in dynamic ECG signal analysis and R wave extraction, and can basically reach more than 94%.
【学位授予单位】：东北大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：R318.04

【参考文献】