Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Identification of cardiovascular diseases through ECG classification using wavelet transformation

Morteza Maleki, Foad Haeri

TL;DR

The paper addresses automatic identification of cardiovascular diseases from ECG signals using wavelet-based feature extraction. It compares continuous and discrete wavelet transforms to extract multiscale features from the MIT-BIH Arrhythmia Database and evaluates multiple classifiers. Random Forest and Gradient Boosting achieve the highest test accuracy (~0.96), with sym5 providing effective time-frequency localization. The findings support integrating wavelet-derived features into automated ECG diagnostics, while highlighting the need for feature selection and regularization to mitigate overfitting and improve clinical deployment.

Abstract

Cardiovascular diseases are the leading cause of mortality globally, necessitating advancements in diagnostic techniques. This study explores the application of wavelet transformation for classifying electrocardiogram (ECG) signals to identify various cardiovascular conditions. Utilizing the MIT-BIH Arrhythmia Database, we employed both continuous and discrete wavelet transforms to decompose ECG signals into frequency sub-bands, from which we extracted eight statistical features per band. These features were then used to train and test various classifiers, including K-Nearest Neighbors and Support Vector Machines, among others. The classifiers demonstrated high efficacy, with some achieving an accuracy of up to 96% on test data, suggesting that wavelet-based feature extraction significantly enhances the prediction of cardiovascular abnormalities in ECG data. The findings advocate for further exploration of wavelet transforms in medical diagnostics to improve automation and accuracy in disease detection. Future work will focus on optimizing feature selection and classifier parameters to refine predictive performance further.

Identification of cardiovascular diseases through ECG classification using wavelet transformation

TL;DR

The paper addresses automatic identification of cardiovascular diseases from ECG signals using wavelet-based feature extraction. It compares continuous and discrete wavelet transforms to extract multiscale features from the MIT-BIH Arrhythmia Database and evaluates multiple classifiers. Random Forest and Gradient Boosting achieve the highest test accuracy (~0.96), with sym5 providing effective time-frequency localization. The findings support integrating wavelet-derived features into automated ECG diagnostics, while highlighting the need for feature selection and regularization to mitigate overfitting and improve clinical deployment.

Abstract

Cardiovascular diseases are the leading cause of mortality globally, necessitating advancements in diagnostic techniques. This study explores the application of wavelet transformation for classifying electrocardiogram (ECG) signals to identify various cardiovascular conditions. Utilizing the MIT-BIH Arrhythmia Database, we employed both continuous and discrete wavelet transforms to decompose ECG signals into frequency sub-bands, from which we extracted eight statistical features per band. These features were then used to train and test various classifiers, including K-Nearest Neighbors and Support Vector Machines, among others. The classifiers demonstrated high efficacy, with some achieving an accuracy of up to 96% on test data, suggesting that wavelet-based feature extraction significantly enhances the prediction of cardiovascular abnormalities in ECG data. The findings advocate for further exploration of wavelet transforms in medical diagnostics to improve automation and accuracy in disease detection. Future work will focus on optimizing feature selection and classifier parameters to refine predictive performance further.
Paper Structure (15 sections, 2 equations, 6 figures, 1 table)

This paper contains 15 sections, 2 equations, 6 figures, 1 table.

Table of Contents

  1. Introduction
  2. Literature Review
  3. Material and Methods
  4. Data Sources
  5. Wavelet Transformation
  6. Feature Extraction
  7. Classification Models
  8. Implementation Details
  9. Results
  10. Wavelet Choice and Decomposition
  11. Feature Extraction and Dataset Preparation
  12. Classifier Performance and Evaluation
  13. Analysis of Classifier Overfitting and Model Complexity
  14. Significance of Results
  15. Discussion and Conclusions

Figures (6)

  • Figure 1: Difference between a sine-wave and wavelet used by Fourier and wavelet transform respectively
  • Figure 2: Different wavelets can be applied for decomposition based on the type of a signal
  • Figure 3: Wavelet transformation decomposes a signal into a set of wavelets of different scales and positions
  • Figure 4: Schematic of the classification approach, in which the outputs of DWT function are used to extract statistical features and incorporate them as inputs into the classifier. The original schema is reported in Taspinar2018 article, and utilized here with some modifications.
  • Figure 5: Typical shape of sym5 wavelet used to decompose the ECG signals.
  • ...and 1 more figures