基于自适应多元多尺度色散熵的心电信号分类研究_毕业论文.pdf

云南大学（专业）硕士学位论文

II

关键词：心电信号分类；联合降噪；正弦辅助；多元经验模态分解；自适应；多元

多尺度色散熵；

Abstract

III

Abstract

With the accelerated pace of work and changing lifestyles, the number of people

suffering from heart-related diseases is hitting record highs. As one of the effective means

studied to provide intuitive classification results, aiming to provide clinical guidance for

doctors, improve the diagnostic accuracy and reduce misdiagnosis.

Affected by objective factors such as complex human tissue, various types of ECGs,

and weak amplitudes, the identification of ECG signals is faced with great challenges. At

present, the research on ECG signals mainly focuses on single-channel data, and ECG

signal acquisition is easily affected by the outside world. interference, the volatility of the

data will greatly reduce the reliability of the data. In view of the above situation, this

thesis takes advantage of the advantages of multi-channel multivariate data, and proposes

a research scheme of ECG signal classification based on adaptive multivariate multiscale

dispersion entropy. The main research content as follows:

ECG signal, and then the proposed a joint denoising scheme based on empirical mode

decomposition to filter out EMG interference in ECG signals.

(3)The sine-assisted multivariate empirical modal decomposition algorithm is used

云南大学（专业）硕士学位论文

IV

to decompose the multivariate data of the multi-channel ECG signal, and the

comprehensiveness of the signal characteristics, the stability and efficiency of the

dispersion entropy are measured in combination with the multi-scale, so as to accumulate

the multivariate modes. Instead of the traditional coarse-grained sampling, an adaptive

multi-scale multiscale dispersion entropy feature extraction scheme is proposed, which

avoids parameter setting, ensures the robustness of the feature entropy value, and

improves the subsequent classification accuracy.

the proposed scheme is tested and analyzed by using the simulated data and the MIT-BIH

data set collected by MIT. The results show that the research on ECG signal classification

in this thesis can provide clinical arrhythmia identification and detection provides

practical ideas.

mode decomposition；Joint denoising；Adaptive；Multivariate multiscale dispersion

entropy

摘 要 ................................................................................................................................. I

Abstract .......................................................................................................................... III

第一章 绪论 .................................................................................................................... 1

第二章 心电信号产生机理 ............................................................................................ 9

2.1 心电信号产生机理及信号特性 ........................................................................ 9

2.2 心电干扰及失常 ............................................................................................. 11

3.3 提出的联合降噪模型 ..................................................................................... 20

云南大学（专业）硕士学位论文

VI

第四章 基于自适应多元多尺度色散熵的特征提取 ................................................... 27

4.3.1 模拟实验 ............................................................................................... 29

4.3.2 数据集实验 ........................................................................................... 31

第五章 心电信号的分类识别 ....................................................................................... 39

5.1.1 线性支持向量机 ................................................................................... 39

5.1.2 非线性支持向量机 ............................................................................... 41