动态脉搏信号检测与脉率变异性实时分析方法研究

发布时间：2018-04-01 10:25

  本文选题：动态脉搏信号　切入点：动态脉率变异性　出处：《兰州理工大学》2015年博士论文

【摘要】：脉搏信号和脉率变异性信号中蕴含着丰富的有关心血管系统的生理和病理信息,可用于心血管疾病的预防、诊断和治疗。脉搏信号易于采集,已成为便携式医疗设备的常用信号之一。检测动态脉搏信号,从中提取并实时分析脉率变异性信号,提取相关信息,对心血管疾病的在线监护与预警具有重要意义。本文介绍了脉搏信号和脉率变异性信号的生理机理和临床价值,综述了脉搏信号检测系统与检测方法及脉率变异性提取与分析方法的国内外研究现状。在已有的研究基础上,针对脉搏信号检测过程中噪声和干扰抑制及信号干扰段检测和质量评估问题,以及现有脉率变异性信号提取和分析方法准确性和实时性不能兼顾的矛盾,提出了相应的解决方法,并将其用于动态脉搏信号检测及脉率变异性信号实时提取与分析。论文主要的研究成果为:1)对于动态脉搏信号中一些可抑制的噪声和干扰,提出了自调参数整系数滤波法,用于不同信噪比的动态脉搏信号滤波。在整系数滤波器的基础上,定义信号光滑度,用于评价滤波结果及自行调节滤波器参数。将所提出方法用于仿真和实测脉搏信号,实验结果表明,相比于常用的滤波方法,所提出方法可快速有效地抑制不同信噪比动态脉搏信号中可抑制噪声和干扰。2)对动态脉搏信号中一些不能通过滤波方法抑制的干扰段,提出干扰段分类检测法,并根据干扰段检测结果进行信号质量评估。根据干扰段的特点将其分为脉冲干扰、脉搏信号丢失段和运动伪迹,提出相应的检测方法,即干扰段分类检测法。通过国际公认数据库及课题组实测脉搏信号验证,结果表明,相比于常用的干扰段检测方法,所提出的方法可快速准确地检测出脉搏信号中的干扰段。同时,根据干扰段检测结果,对信号进行质量评估。去除质量低的信号段,为进一步动态脉搏信号处理提供信号质量保证。3)根据脉搏信号时域和频域特征,总结常用脉率变异性信号提取方法的优缺点,提出了自适应幅度阈值法、基于改进滑窗迭代DFT的脉率变异性提取法和基于Hilbert-Huang变换的脉率变异性提取法。将所提出的方法用于仿真和实测脉搏信号,实验结果表明,相比于其它方法,滑窗迭代DFT(基波)法可准确实时地提取脉率变异性信号,且对脉搏信号的信噪比和采样频率变化不敏感,可用于动态脉率变异性信号实时提取。4)对于脉率变异性信号的一些时域和非线性分析方法存在的实时性不高、算法运算量大等不足,采用滑窗迭代思想对其改进,并就改进后方法是否用于心血管疾病识别进行了探索。对时域法、庞加莱散点图法、基本尺度熵分析法和符号序列熵分析法进行改进,并将其用于实测脉率变异性信号分析。结果表明,相比之下,改进后方法可在脉率变异性信号数据点更新的同时,快速对其进行分析,提取相应的参数,实时性得到质的提高,可用于动态脉率变异性信号实时分析。将改进方法用于分析国际公认数据中的年轻人和老年人、健康人和冠心病人的脉率变异性信号,提取一些参数组成特征向量,用于智能学习方法分类。相比之下,分类的准确率很高,表明所改进的方法可用于一些心血管疾病的识别。5)基于智能手机平台,研制了动态脉搏信号检测与处理系统,验证所提出方法实用性。通过实际检测和处理动态脉搏信号,结果表明:所提出的自调参数整系数滤波法、基于干扰段分类检测的信号质量评估法可用于动态脉搏信号检测;所提出的基于改进滑窗迭代DFT(基波)法,所改进的脉率变异性信号分析法可准确实时地提取并分析动态脉率变异性信号,提取相关的参数。为进一步实现心血管疾病在线监护与预警提供保障。
[Abstract]:The pulse signal and the pulse rate variability signal contained in the physiological and pathological information about the cardiovascular system of the rich, can be used for cardiovascular disease prevention, diagnosis and treatment. The pulse signal is easy to be collected, has become one of the most commonly used signal portable medical equipment. The detection of dynamic pulse signal, extracted from real-time analysis and pulse rate variability signal and extract relevant information, plays an important role in online monitoring and early warning of cardiovascular diseases. This paper introduces the clinical value and physiological mechanism of pulse signal and pulse rate variability signals, the research status of pulse signal detection system and detection method and the method of extraction and analysis of pulse rate variability was reviewed based on the existing research. On the pulse signal detection in noise and interference suppression and signal interference detection and quality evaluation, and the current pulse rate variability signal extraction and Analysis of the contradiction between accuracy and real-time method can not take into account, puts forward corresponding solving methods, and applied to real time extraction and analysis of variability of dynamic signal pulse signal detection and pulse rate. The main research results are as follows: 1) for some can suppress noise and interference of pulse signal, the self adjusting parameter integer coefficient filter method for dynamic pulse signal filtering different SNR. Based on integer coefficient filter, defined signal smoothness, and adjust the filtering results for the evaluation of the filter parameters. The proposed method is used for the simulation and the measured pulse signal, the experimental results show that compared with common filter methods, the proposed the method can effectively suppress the different SNR of dynamic pulse signal can suppress noise and interference.2) of interference section cannot be suppressed by dynamic pulse signal filtering method is proposed. Interference segment classification detection method, and signal quality assessment based on the interference section of test results. According to the characteristics of interference section will be divided into pulse interference, pulse signal loss and motion artifacts, the corresponding detection methods, namely interference segment classification method. Through the internationally recognized data base and research group measured pulse signal verification. The results show that, compared to the interference detection method used, the proposed method can detect the interference section of the pulse signal quickly and accurately. At the same time, according to the interference period of test results, to evaluate the quality of the signal. The signal segment removal of low quality, provide quality assurance for the further dynamic.3 signal processing of pulse signal according to the pulse) signal time domain and frequency domain feature extraction method, summarize the advantages and disadvantages of common pulse rate variability signal, the adaptive amplitude threshold method, the improved sliding window iterative DFT extraction based on pulse rate variability And the method based on Hilbert-Huang transform pulse rate variability extraction method. The proposed method is applied to the simulation and measurement of pulse signal, the experimental results show that compared with other methods, sliding window iterative DFT (Ji Bo) accurate real-time extraction of pulse rate variability signals, and the signal-to-noise ratio and the sampling frequency is not sensitive to the change of the pulse signal, can be used for dynamic pulse rate variability signal extraction.4) for real-time presence of pulse rate variability signals time domain and the nonlinear analysis method is not high, lack of a large amount of computation, using a sliding window iterative thought to its improvement, and the improved method is used to identify the cardiovascular disease exploration on. Time domain method, Poincare plot method, basic scale entropy analysis and sequence entropy analysis method was improved and used for the analysis of measured pulse rate variability signals. The results show that, compared with the improved Method can be used in pulse rate variability signal data update at the same time, the rapid analysis of the extraction of the corresponding parameters, real-time quality improvement, can be used for dynamic pulse rate variability signal analysis. The improved method is used to analyze the data of the internationally recognized young people and older people, healthy people Wacom heart patients the pulse rate variability signal, extracting some parameters of feature vector for intelligent learning method classification. By contrast, the classification accuracy is very high, show that the improved method can be used for the identification of some cardiovascular diseases).5 intelligent mobile phone platform is developed based on dynamic pulse signal detection and processing system, verify the applicability of the proposed methods. Through the actual detection and processing of dynamic pulse signal, the results show that the self adjusting parameter coefficient filtering method is proposed, based on the signal quality assessment method of interference detection can be used for dynamic segment classification Pulse signal detection; improved sliding window iterative algorithm of DFT based on the proposed (Ji Bo) method, the improved pulse rate variability signal analysis can be accurately extracted in real time and dynamic analysis of pulse rate variability signals, extract the relevant parameters. For the further implementation of online monitoring and early warning of cardiovascular disease to provide protection.

【学位授予单位】：兰州理工大学
【学位级别】：博士
【学位授予年份】：2015
【分类号】：TN911.23;R540.4
，

本文编号：1695179