Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Deep Neural Network Architectures for Electrocardiogram Classification: A Comprehensive Evaluation

Yun Song, Wenjia Zheng, Tiedan Chen, Ziyu Wang, Jiazhao Shi, Yisong Chen

TL;DR

It is demonstrated that leveraging complementary deep architectures significantly enhances classification reliability, providing a robust and interpretable foundation for intelligent arrhythmia detection systems.

Abstract

With the rising prevalence of cardiovascular diseases, electrocardiograms (ECG) remain essential for the non-invasive detection of cardiac abnormalities. This study presents a comprehensive evaluation of deep neural network architectures for automated arrhythmia classification, integrating temporal modeling, attention mechanisms, and ensemble strategies. To address data scarcity in minority classes, the MIT-BIH Arrhythmia dataset was augmented using a Generative Adversarial Network (GAN). We developed and compared four distinct architectures, including Convolutional Neural Networks (CNN), CNN combined with Long Short-Term Memory (CNN-LSTM), CNN-LSTM with Attention, and 1D Residual Networks (ResNet-1D), to capture both local morphological features and long-term temporal dependencies. Performance was rigorously evaluated using accuracy, F1-score, and Area Under the Curve (AUC) with 95\% confidence intervals to ensure statistical robustness, while Gradient-weighted Class Activation Mapping (Grad-CAM) was employed to validate model interpretability. Experimental results indicate that the CNN-LSTM model achieved the optimal stand-alone balance between sensitivity and specificity, yielding an F1-score of 0.951. Conversely, the CNN-LSTM-Attention and ResNet-1D models exhibited higher sensitivity to class imbalance. To mitigate this, a dynamic ensemble fusion strategy was introduced; specifically, the Top2-Weighted ensemble achieved the highest overall performance with an F1-score of 0.958. These findings demonstrate that leveraging complementary deep architectures significantly enhances classification reliability, providing a robust and interpretable foundation for intelligent arrhythmia detection systems.

Deep Neural Network Architectures for Electrocardiogram Classification: A Comprehensive Evaluation

TL;DR

It is demonstrated that leveraging complementary deep architectures significantly enhances classification reliability, providing a robust and interpretable foundation for intelligent arrhythmia detection systems.

Abstract

With the rising prevalence of cardiovascular diseases, electrocardiograms (ECG) remain essential for the non-invasive detection of cardiac abnormalities. This study presents a comprehensive evaluation of deep neural network architectures for automated arrhythmia classification, integrating temporal modeling, attention mechanisms, and ensemble strategies. To address data scarcity in minority classes, the MIT-BIH Arrhythmia dataset was augmented using a Generative Adversarial Network (GAN). We developed and compared four distinct architectures, including Convolutional Neural Networks (CNN), CNN combined with Long Short-Term Memory (CNN-LSTM), CNN-LSTM with Attention, and 1D Residual Networks (ResNet-1D), to capture both local morphological features and long-term temporal dependencies. Performance was rigorously evaluated using accuracy, F1-score, and Area Under the Curve (AUC) with 95\% confidence intervals to ensure statistical robustness, while Gradient-weighted Class Activation Mapping (Grad-CAM) was employed to validate model interpretability. Experimental results indicate that the CNN-LSTM model achieved the optimal stand-alone balance between sensitivity and specificity, yielding an F1-score of 0.951. Conversely, the CNN-LSTM-Attention and ResNet-1D models exhibited higher sensitivity to class imbalance. To mitigate this, a dynamic ensemble fusion strategy was introduced; specifically, the Top2-Weighted ensemble achieved the highest overall performance with an F1-score of 0.958. These findings demonstrate that leveraging complementary deep architectures significantly enhances classification reliability, providing a robust and interpretable foundation for intelligent arrhythmia detection systems.
Paper Structure (40 sections, 14 equations, 12 figures, 2 tables)

This paper contains 40 sections, 14 equations, 12 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Methodology
  4. Dataset Preprocessing for ECG Analysis
  5. Dataset and Preprocessing
  6. Dataset Overview
  7. Category Selection and Label Mapping
  8. Data Preprocessing Pipeline
  9. Heartbeat Segmentation
  10. Label Encoding
  11. Signal Normalization
  12. Dataset Partitioning
  13. Data Augmentation to Balance Data
  14. ECG Waveform Visualization
  15. Model Architectures
  16. ...and 25 more sections

Figures (12)

  • Figure 1: Time-domain waveform examples of the five heartbeat classes: Normal (N), Atrial Premature (A), Premature Ventricular (V), Fusion of Paced and Normal (f), and Fusion of Ventricular and Normal (F).
  • Figure 2: CNN Model Architecture
  • Figure 3: CNN-LSTM Model Architecture
  • Figure 4: CNN-LSTM-Attention Model Architecture
  • Figure 5: ResNet-1D Model Architecture
  • ...and 7 more figures