发布时间：2015-01-14 来源：人大经济论坛

电子信息工程毕业论文范文 目录 摘要 I - ABSTRACTII - 第一章：绪论1 1.1 课题研究背景介绍1 1.2 课题研究的意义1 第二章：课题研究背景知识3 2.1 心脏的基本结构：3 2.2 心电图4 2.3 心律失常及ECG特征6 2.3.1 正常窦性心律（NSR）6 2.3.2 心室性心动过速（VT）7 2.3.3 心室纤维性颤动（VF）8 第三章:方法10 3.1 ECG信号特征提取方法10 3.1.1 AR建模法10 3.1.2 建模阶次确定11 3.2 分类的技术12 第四章：系统的实现15 4.1 AR建模15 4.2 基于普通线性分类器的分类15 第五章：结果与讨论18 5.1 基于线性分类器的分类结果18 致谢21 参考文献22 附录1 程序24 摘要 本文的基本思想是从病人的ECG信号中采用数据处理手段自动提取与病况有关的特征，再应用线性分类技术将病人的ECG信号进行自动分类，帮助临床医生快速、准确地进行心脏病症的诊断。 本文以MIT-BIH数据库中的正常窦性心律、心室性心动过速和心室纤维性颤动信号为研究对象，提取信号的特征并进行分类。具体的方法是：利用单变量回归（AR）模型对正常窦性心律、心室性心动过速和心室性纤维性颤动信号进行数学建模；给出一种提取心电信号特征信息的方法，即以AR模型系数为信号的特征；再利用线性分类器进行分类和比较。 在根据实验结果得到：利用心电信号对心率失常进行计算机自动诊断的可行方法。 关键词：心电图, 心率失常, 特征值, 分类, 诊断 ABSTRACT The main method of this paper is based on obtaining the signal data from patient ECG signal, processing the data, then extract the feature of the ECGs. After that, using Linear classification technique was used for the automatic detection, which can help doctors make the diagnosis accurately and quickly. Normal sinus rhythm, ventricular tachycardia and ventricular fibrillation signals which collect from MIT-BIH database were used for the analysis in current research including Autoregressive (AR) modeling, feature extraction and classification. The AR modeling was applied to normal sinus rhythm, ventricular tachycardia and ventricular fibrillation signals. AR model coefficients were utilized as the ECG features for the classification and diagnosis, and linear classifier was employed to classify the features. Finally, the conclusion can be drown: It is feasible and practical that using ECG signal can diagnose the arrhythmia automatically through computers automatically. Keywords: ECGs, arrhythmia, features, classification, diagnosis