Multistage Arrhythmia Classification using Dual-Tree Complex Wavelet Transform and Hybrid Deep Learning Models

DOI

10.2991/978-94-6463-866-0_60How to use a DOI?

Keywords

Dual-Tree Complex Wavelet Transform; Hybrid CNN-Transformer Model; DenseNet-121 Architecture; Cardiac Arrhythmia Detection; Ventricular Tachycardia; Ventricular Fibrillation; Deep Learning for ECG; Implantable Defibrillator Support; Vision Transformer (ViT); Biomedical Feature Extraction

Abstract

Accurate and precise classification of cardiac arrhythmias is essential for early diagnosis and prevention of life-threatening cardiac conditions. This research introduces an innovative two-stage deep learning approach for classifying arrhythmias, utilizing Dual-Tree Complex Wavelet Transform (DT-CWT) for feature extraction, followed by a hybrid CNN-Transformer model and DenseNet-121 for layered classification. The DT-CWT effectively captures both time and frequency-domain characteristics of ECG signals, providing the richer feature representations for improved classification accuracy. In the first stage, a Hybrid CNN-Transformer (CNN + ViT) distinguishes between Shockable Arrhythmia (SA) and Non-Shockable Arrhythmia (NSA). In the second stage, DenseNet further classifies the detected arrhythmias into specific types as Ventricular Tachycardia (VT) and Ventricular Fibrillation (VF). Experimental validation is conducted on benchmark ECG datasets, and results demonstrate superior classification performance compared to existing approaches, achieving high Accuracy, F1-Score, and AUC-ROC. The proposed method enhances automatic ECG classification, ensuring improved reliability and clinical applicability in real-world settings.

Copyright

© 2025 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - K. M. Bharath
AU  - M. Kowsigan
PY  - 2025
DA  - 2025/10/31
TI  - Multistage Arrhythmia Classification using Dual-Tree Complex Wavelet Transform and Hybrid Deep Learning Models
BT  - Proceedings of the International Conference on Intelligent Systems and Digital Transformation (ICISD 2025)
PB  - Atlantis Press
SP  - 728
EP  - 746
SN  - 2589-4919
UR  - https://doi.org/10.2991/978-94-6463-866-0_60
DO  - 10.2991/978-94-6463-866-0_60
ID  - Bharath2025
ER  -