基于深度学习和模糊C均值的心电信号分类方法

被引：56

作者：

吴志勇 ^{[1
,2
]}

丁香乾 ^{[1
]}

许晓伟 ^{[1
]}

鞠传香 ^{[2
]}

机构：

[1] 中国海洋大学信息科学与工程学院

[2] 山东理工大学计算机科学与技术学院

来源：

自动化学报 | 2018年 / 44卷 / 10期

基金：

国家重点研发计划;

关键词：

心电信号分类; 深度学习; 模糊C均值; 深度信念网络;

D O I：

暂无

中图分类号：

TN911.7 [信号处理]; R540.4 [诊断学];

学科分类号：

081002 [信号与信息处理]; 100201 [内科学];

摘要：

针对长时海量心电信号自动分类系统中,心电专家诊断费时、费力和成本高,心电信号形态复杂导致特征提取困难,异常诊断模型适应性差、准确度低等问题,本文提出一种基于深度学习和模糊C均值的心电信号分类方法.该方法主要包括心电信号降噪预处理、心电信号分段和采样点统一化、无监督心跳特征学习、模糊C均值分类4个步骤,给出了模糊C均值深度信念网络FCMDBN模型结构和学习分类算法.仿真实验基于MIT-BIH心率异常数据库表明,与基于传统心电特征人工设计的分类方法相比,本文提出的信号诊断方法具有较高的适应性和准确度.

引用

页码：1913 / 1920

页数：8

共 28 条

[1]

Real-Time Patient-Specific ECG Classification by 1-D Convolutional Neural Networks.[J] Kiranyaz Serkan;Ince Turker;Gabbouj Moncef IEEE transactions on bio-medical engineering 2016,

[2]

Deep learning approach for active classification of electrocardiogram signals[J] M.M. Al Rahhal;Yakoub Bazi;Haikel AlHichri;Naif Alajlan;Farid Melgani;R.R. Yager Information Sciences 2016,

[3]

ECG heart beat classification method based on modified ABC algorithm[J] Selim Dilmac;Mehmet Korurek Applied Soft Computing 2015,

[4]

Auto-encoder based dimensionality reduction[J] Yasi Wang;Hongxun Yao;Sicheng Zhao Neurocomputing 2016,

[5]

Deep Reconstruction Models for Image Set Classification.[J] Hayat Munawar;Bennamoun Mohammed;An Senjian IEEE transactions on pattern analysis and machine intelligence 2015,

[6]

Feature Extraction of Electrocardiogram Signals by Applying Adaptive Threshold and Principal Component Analysis[J] R. Rodríguez;A. Mexicano;J. Bila;S. Cervantes;R. Ponce Journal of Applied Research and Technology 2015,

[7]

Deep learning in neural networks: An overview[J] Jürgen Schmidhuber Neural Networks 2015,

[8]

ECG beat classification using PCA; LDA; ICA and Discrete Wavelet Transform[J] Roshan Joy Martis;U. Rajendra Acharya;Lim Choo Min Biomedical Signal Processing and Control 2013,

[9]

ECG arrhythmia classification based on optimum-path forest[J] Expert Systems With Applications 2012,

[10]

Classification of Parkinson's disease using feature weighting method on the basis of fuzzy C-means clustering[J] Kemal Polat International Journal of Systems Science 2012,

← 1 2 3 →