A Novel Deep 2D-CNN Model for ECG-Based Arrhythmia Diagnosis with Selective Attention Mechanism and CWT Integration

A Novel Deep 2D-CNN Model for ECG-Based Arrhythmia Diagnosis with Selective Attention Mechanism and CWT Integration

This study introduces an innovative approach for arrhythmia diagnosis via electrocardiogram (ECG) signals, employing a 2D Convolutional Neural Network (CNN) model fused with a Continuous Wavelet Transform (CWT) and a Selective Attention Mechanism (SAM). The SAM enhances feature focus, improving clas...

Full description

Saved in:
Bibliographic Details
Main Authors: Hassanain Shakir Mansour, Morteza Valizadeh, Alaa Hussein Abdulaal, Mehdi Chehel Amirani
Format: Article
Language:English
Published: Faculty of Engineering, University of Kufa 2025-04-01
Series:Mağallaẗ Al-kūfaẗ Al-handasiyyaẗ
Subjects:
arrhythmia
ecg signals
cnn
continuous wavelet transform
selective attention mechanism
Online Access:https://journal.uokufa.edu.iq/index.php/kje/article/view/17579
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:This study introduces an innovative approach for arrhythmia diagnosis via electrocardiogram (ECG) signals, employing a 2D Convolutional Neural Network (CNN) model fused with a Continuous Wavelet Transform (CWT) and a Selective Attention Mechanism (SAM). The SAM enhances feature focus, improving classification accuracy. The model effectively categorizes ECG signals into Normal and Abnormal classes, subcategorizing Abnormal patterns into four types. Merging deep learning with signal processing enables the correct classification of arrhythmia. The outcome yields high accuracy, such as 99.784% for multi-class and 99.94% for binary classification. The methodology proposed in this paper shows the capability of secure yielding for arrhythmia diagnosis, setting novel standards within healthcare applications.
ISSN:2071-5528
2523-0018

Similar Items