ECG Classification System for Arrhythmia Detection Using Convolutional Neural Networks

TL;DR

The paper presents a CNN-based ECG beat classification system with a residual block to detect five arrhythmia categories (NOR, LBBB, RBBB, PVC, APC) using MIT-BIH data. Preprocessing combines a moving-average filter with eight-level Daubechies-4 wavelet denoising, producing 180-sample beat segments that feed directly into an end-to-end CNN. On lead II data, the model achieves high performance (approximately 97.8% accuracy, 97% sensitivity, 97.32% specificity) with a 50/50 train/test split after a 300-epoch training regime. The work demonstrates clinical viability for fast, automated ECG interpretation and outlines future directions to employ deeper residual architectures to push performance further.

Abstract

Arrhythmia is just one of the many cardiovascular illnesses that have been extensively studied throughout the years. Using multi-lead ECG data, this research describes a deep learning (DL) pipeline technique based on convolutional neural network (CNN) algorithms to detect cardiovascular lar arrhythmia in patients. The suggested model architecture has hidden layers with a residual block in addition to the input and output layers. In this study, the classification of the ECG signals into five main groups, namely: Left Bundle Branch Block (LBBB), Right Bundle Branch Block (RBBB), Atrial Premature Contraction (APC), Premature Ventricular Contraction (PVC), and Normal Beat (N), are performed. Using the MIT-BIH arrhythmia dataset, we assessed the suggested technique. The findings show that our suggested strategy classified 15,000 cases with a high accuracy of 98.2%

Figures (12)

• Figure 1: Model Pipeline
• Figure 2: Overview of the ETL process
• Figure 3: Dataset example
• Figure 4: Input Signal Data
• Figure 5: Class imbalance rebalancing