Classifying Heart Sounds Using Images of Motifs, MFCC and Temporal Features

被引:0
|
作者
Diogo Marcelo Nogueira
Carlos Abreu Ferreira
Elsa Ferreira Gomes
Alípio M. Jorge
机构
[1] INESC TEC,
[2] Instituto Superior de Engenharia do Porto,undefined
[3] Faculdade de Ciências da Universidade do Porto,undefined
来源
Journal of Medical Systems | 2019年 / 43卷
关键词
Phonocardiogram; Electrocardiogram; Mel-frequency cepstral coefficients; Motifs; Time features;
D O I
暂无
中图分类号
学科分类号
摘要
Cardiovascular disease is the leading cause of death in the world, and its early detection is a key to improving long-term health outcomes. The auscultation of the heart is still an important method in the medical process because it is very simple and cheap. To detect possible heart anomalies at an early stage, an automatic method enabling cardiac health low-cost screening for the general population would be highly valuable. By analyzing the phonocardiogram signals, it is possible to perform cardiac diagnosis and find possible anomalies at an early-term. Therefore, the development of intelligent and automated analysis tools of the phonocardiogram is very relevant. In this work, we use simultaneously collected electrocardiograms and phonocardiograms from the Physionet Challenge database with the main objective of determining whether a phonocardiogram corresponds to a “normal” or “abnormal” physiological state. Our main contribution is the methodological combination of time domain features and frequency domain features of phonocardiogram signals to improve cardiac disease automatic classification. This novel approach is developed using both features. First, the phonocardiogram signals are segmented with an algorithm based on a logistic regression hidden semi-Markov model, which uses electrocardiogram signals as a reference. Then, two groups of features from the time and frequency domain are extracted from the phonocardiogram segments. One group is based on motifs and the other on Mel-frequency cepstral coefficients. After that, we combine these features into a two-dimensional time-frequency heat map representation. Lastly, a binary classifier is applied to both groups of features to learn a model that discriminates between normal and abnormal phonocardiogram signals. In the experiments, three classification algorithms are used: Support Vector Machines, Convolutional Neural Network, and Random Forest. The best results are achieved when both time and Mel-frequency cepstral coefficients features are considered using a Support Vector Machines with a radial kernel.
引用
收藏
相关论文
共 50 条
  • [31] Detecting age-related macular degeneration (AMD) biomarker images using MFCC and texture features
    Wang, Yiyang
    Ma, Xufan
    Weddell, Rob
    Okemgbo, Abum
    Rein, David
    Fawzi, Amani A.
    Furst, Jacob
    Raicu, Daniela
    MEDICAL IMAGING 2020: COMPUTER-AIDED DIAGNOSIS, 2020, 11314
  • [32] Classifying environmental sounds using image recognition networks
    Boddapati, Venkatesh
    Petef, Andrej
    Rasmusson, Jim
    Lundberg, Lars
    KNOWLEDGE-BASED AND INTELLIGENT INFORMATION & ENGINEERING SYSTEMS, 2017, 112 : 2048 - 2056
  • [33] Phonocardiogram signals processing approach for PASCAL Classifying Heart Sounds Challenge
    Chakir, Fatima
    Jilbab, Abdelilah
    Nacir, Chafik
    Hammouch, Ahmed
    SIGNAL IMAGE AND VIDEO PROCESSING, 2018, 12 (06) : 1149 - 1155
  • [34] Phonocardiogram signals processing approach for PASCAL Classifying Heart Sounds Challenge
    Fatima Chakir
    Abdelilah Jilbab
    Chafik Nacir
    Ahmed Hammouch
    Signal, Image and Video Processing, 2018, 12 : 1149 - 1155
  • [35] Using random walker for knowledge transfer in classifying multi-temporal VHR images
    Guo, Zhou
    Du, Shihong
    Zhao, Wenzhi
    INTERNATIONAL JOURNAL OF REMOTE SENSING, 2015, 36 (17) : 4332 - 4343
  • [36] A comparative approach between cepstral features for human authentication using heart sounds
    Abo-Zahhad, M.
    Farrag, Mohammed
    Abbas, Sherif N.
    Ahmed, Sabah M.
    SIGNAL IMAGE AND VIDEO PROCESSING, 2016, 10 (05) : 843 - 851
  • [37] A comparative approach between cepstral features for human authentication using heart sounds
    M. Abo-Zahhad
    Mohammed Farrag
    Sherif N. Abbas
    Sabah M. Ahmed
    Signal, Image and Video Processing, 2016, 10 : 843 - 851
  • [38] FireNet-MLstm for classifying liver lesions by using deep features in CT images
    Gedeon Kashala Kabe
    Yuqing Song
    Zhe Liu
    Multimedia Tools and Applications, 2022, 81 : 1607 - 1623
  • [39] Source Microphone Identification Using Multitaper MFCC Features
    Eskidere, Omer
    Karatutlu, Ali
    2015 9TH INTERNATIONAL CONFERENCE ON ELECTRICAL AND ELECTRONICS ENGINEERING (ELECO), 2015, : 227 - 231
  • [40] FireNet-MLstm for classifying liver lesions by using deep features in CT images
    Kabe, Gedeon Kashala
    Song, Yuqing
    Liu, Zhe
    MULTIMEDIA TOOLS AND APPLICATIONS, 2022, 81 (02) : 1607 - 1623