心电分析系统的设计与实现

发布时间：2018-08-07 11:09

【摘要】：近年来,随着青年人猝死的现象越来越多,心脏的健康问题再一次被人们谈及并加以重视,促使出现可穿戴设备骤然繁荣的现象。其实,心脏系统方面的疾病历年来都是医学领域特别重视的主题,其具有疾病种类繁多、隐匿性高、致死率高等特点,对于人类生命健康构成极大的安全威胁。因此,对于心电信号的研究是十分有必要意义的,其中研究包括对采集的心电信号进行预处理、心脏特征波形识别检测以及通过识别的特征波形对心律失常相关疾病加以分类诊断的研究。心电信号预处理的结果决定着心电波形的检测准确效果,波形检测结果直接影响对心律失常相关疾病的分类与诊断效果。针对上述问题,本文主要围绕着心电信号预处理、心电波形特征检测以及心律失常疾病分类诊断这三个方面进行阐述。首先,针对心电信号数据文件的特点进行数据读取算法的设计与实现,然后根据心电信号噪声特点,采用多尺度小波分解与重构的方法设计心电信号去噪算法,并对去噪效果与传统的滤波器去噪法进行对比分析。其次,对去噪后的心电信号进行波形识别,根据心电波形的特点,设计并实现了一种非线性放大R波检测算法。此外,针对心电波形的区域位置特点,设计并实现了一种新颖快速的R波检测算法,通过构建脉冲波与区域定位特点进行检测,并提出了相应的漏检与误检补充算法并对R波检测结果进行验证与对比。基于构建脉冲波与区域定位的思想对Q波、S波、QRS波起止位置点以及P波、T波、P波起止位置点、T波起止位置点进行检测算法的设计与实现并对各个心电波形的检测结果进行分析。最后,对心电波形检测结果提取心电特征参数,在临床医学诊断知识的基础下构建规则决策树对几种心律失常疾病进行分类与诊断,并对算法实现的结果进行分析。本文在以上三个方面研究内容的基础上,应用Windows下Qt软件平台对心电分析系统进行设计与实现。通过对软件的运行和结果展示,该软件运行流畅无错误,对心电信号预处理效果明显、心电波形检测较为准确、几种心律失常疾病自动分类诊断效果较好,对心电分析工作者有参考价值。
[Abstract]:In recent years, with the phenomenon of sudden death of young people more and more, the problem of heart health has once again been discussed and paid attention to, promoting the appearance of wearable devices suddenly prosperity phenomenon. In fact, the disease of heart system has been paid special attention to in the field of medicine over the years. It has the characteristics of various kinds of diseases, high occult, high fatality rate, which poses a great security threat to human life and health. Therefore, the study of ECG signals is of great significance, including the preprocessing of collected ECG signals. The study of heart characteristic waveform recognition and diagnosis of arrhythmia-related diseases by identifying characteristic waveform. The results of ECG signal preprocessing determine the accuracy of ECG waveform detection, which directly affect the classification and diagnosis of arrhythmia-related diseases. In order to solve the above problems, this paper focuses on three aspects: ECG signal preprocessing, ECG waveform detection and arrhythmia diagnosis. Firstly, according to the characteristics of ECG data file, the data reading algorithm is designed and implemented. Then, according to the characteristics of ECG signal noise, the method of multi-scale wavelet decomposition and reconstruction is used to design ECG signal de-noising algorithm. The denoising effect is compared with the traditional filter denoising method. Secondly, waveform recognition of de-noised ECG signal is carried out. According to the characteristics of ECG waveform, a nonlinear amplification R-wave detection algorithm is designed and implemented. In addition, a novel and fast R-wave detection algorithm is designed and implemented according to the regional location characteristics of ECG waveform. The corresponding algorithms of missing detection and false detection are proposed, and the results of R wave detection are verified and compared. Based on the idea of constructing pulse wave and region location, the algorithm is designed and implemented to detect the starting and ending points of Q wave and S wave QRS wave and the starting and ending points of P wave T wave and P wave. The detection results of each ECG wave are analyzed. Finally, the ECG characteristic parameters are extracted from the ECG waveform detection results. Based on the clinical diagnosis knowledge, the rule decision tree is constructed to classify and diagnose several arrhythmia diseases, and the results of the algorithm are analyzed. Based on the above three aspects, the QT software platform based on Windows is used to design and implement the ECG analysis system. The results show that the software runs smoothly and error-free, has obvious effect on ECG signal preprocessing, accurate ECG waveform detection, and better automatic classification and diagnosis of several arrhythmia diseases. It has reference value for ECG analysts.
【学位授予单位】：哈尔滨工业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：R540.4;TN911.7

【相似文献】