Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Coronary Artery Disease Classification Using One-dimensional Convolutional Neural Network

Atitaya Phoemsuk, Vahid Abolghasemi

TL;DR

This work tackles early, ECG-based CAD detection using a compact 1D-CNN to reduce computational load while maintaining high diagnostic ability. By sourcing data from PhysioNet (MIMIC III, Fantasia) and St. Petersburg, segmenting signals into 8-second windows, and labeling CAD vs non-CAD, the authors evaluate how input length and dropout affect performance. The proposed 12-layer 1D-CNN achieves strong training accuracy (up to 97.3%) with notable unseen-data performance peaks at a 250-sample length (D2: 82.5%, D3: 85.7%), and shows benefits from dropout regularization. Compared to baselines, the model delivers high accuracy with favorable computational metrics (e.g., 65,792 FLOPs) and offers a path toward efficient, real-time CAD screening, though generalization and cross-dataset validation remain areas for further work.

Abstract

Coronary Artery Disease (CAD) diagnostic to be a major global cause of death, necessitating innovative solutions. Addressing the critical importance of early CAD detection and its impact on the mortality rate, we propose the potential of one-dimensional convolutional neural networks (1D-CNN) to enhance detection accuracy and reduce network complexity. This study goes beyond traditional diagnostic methodologies, leveraging the remarkable ability of 1D-CNN to interpret complex patterns within Electrocardiogram (ECG) signals without depending on feature extraction techniques. We explore the impact of varying sample lengths on model performance and conduct experiments involving layers reduction. The ECG data employed were obtained from the PhysioNet databases, namely the MIMIC III and Fantasia datasets, with respective sampling frequencies of 125 Hz and 250 Hz. The highest accuracy for unseen data obtained with a sample length of 250. These initial findings demonstrate the potential of 1D-CNNs in CAD diagnosis using ECG signals and highlight the sample size's role in achieving high accuracy.

Coronary Artery Disease Classification Using One-dimensional Convolutional Neural Network

TL;DR

This work tackles early, ECG-based CAD detection using a compact 1D-CNN to reduce computational load while maintaining high diagnostic ability. By sourcing data from PhysioNet (MIMIC III, Fantasia) and St. Petersburg, segmenting signals into 8-second windows, and labeling CAD vs non-CAD, the authors evaluate how input length and dropout affect performance. The proposed 12-layer 1D-CNN achieves strong training accuracy (up to 97.3%) with notable unseen-data performance peaks at a 250-sample length (D2: 82.5%, D3: 85.7%), and shows benefits from dropout regularization. Compared to baselines, the model delivers high accuracy with favorable computational metrics (e.g., 65,792 FLOPs) and offers a path toward efficient, real-time CAD screening, though generalization and cross-dataset validation remain areas for further work.

Abstract

Coronary Artery Disease (CAD) diagnostic to be a major global cause of death, necessitating innovative solutions. Addressing the critical importance of early CAD detection and its impact on the mortality rate, we propose the potential of one-dimensional convolutional neural networks (1D-CNN) to enhance detection accuracy and reduce network complexity. This study goes beyond traditional diagnostic methodologies, leveraging the remarkable ability of 1D-CNN to interpret complex patterns within Electrocardiogram (ECG) signals without depending on feature extraction techniques. We explore the impact of varying sample lengths on model performance and conduct experiments involving layers reduction. The ECG data employed were obtained from the PhysioNet databases, namely the MIMIC III and Fantasia datasets, with respective sampling frequencies of 125 Hz and 250 Hz. The highest accuracy for unseen data obtained with a sample length of 250. These initial findings demonstrate the potential of 1D-CNNs in CAD diagnosis using ECG signals and highlight the sample size's role in achieving high accuracy.

Paper Structure

This paper contains 8 sections, 2 equations, 3 figures, 4 tables.

Table of Contents

  1. Introduction
  2. Methodology
  3. Data preparation
  4. Data pre-processing
  5. CNN model
  6. Experimental Results
  7. Result
  8. Conclusion

Figures (3)

  • Figure 1: Representation of ECG waveform componentsECGwaves_pic
  • Figure 2: Flowchart of the proposed approach
  • Figure 3: Confusion matrices on dataset $D_2$ with 250-sample length