An electrocardiogram (ECG) is a non-invasive, inexpensive, and widely used diagnostic tool for arrhythmia diagnosis in clinics. Deep learning techniques have shown great promise in ECG signal analysis, enabling automatic and accurate detection of various cardiac arrhythmia. This paper proposes an automated multi-label cardiac arrhythmia classification network based on a convolutional neural network (CNN). The network aims to detect and classify 45 cardiac arrhythmia classes using 12-lead ECG data. Unlike previous studies, our approach incorporates the residual structure and channel attention mechanism. Thus, we developed two key schemes to improve classification performance: the Global Channel Attention Block (GCAB) and the Short Residual Block (SRB). The GCAB incorporates dilated convolutions to preserve overall features. It focuses on the important characteristics of each arrhythmia class from the original electrocardiogram data during the training process. The SRB employs a residual structure to enhance classification accuracy. The network’s performance is evaluated using a large-scale 12-lead electrocardiogram database for arrhythmia study on PhysioNet and the 2018 China Physiological Signal Challenge (CPSC) dataset. In particular, the proposed classification network shows the highest scores in average precision, recall, F1 score, area under the receiver operating characteristic, and accuracy compared to existing CNN-based arrhythmia classification networks in a large-scale 12-lead electrocardiogram database for arrhythmia study on PhysioNet. Finally, to evaluate the performance of the proposed classification network, we compared our proposed network with widely known classification networks such as VGGNet, ResNet, SENet, MobileNet, and EfficientNet. The proposed network demonstrates superior performance compared to other well-known classification networks. We validate the proposed arrhythmia classification network through confusion matrix and AUROC curve.
Alternative Abstract
An electrocardiogram (ECG) is a non-invasive, inexpensive, and widely used diagnostic tool for arrhythmia diagnosis in clinics. Deep learning techniques have shown great promise in ECG signal analysis, enabling automatic and accurate detection of various cardiac arrhythmia. This paper proposes an automated multi-label cardiac arrhythmia classification network based on a convolutional neural network (CNN). The network aims to detect and classify 45 cardiac arrhythmia classes using 12-lead ECG data. Unlike previous studies, our approach incorporates the residual structure and channel attention mechanism. Thus, we developed two key schemes to improve classification performance: the Global Channel Attention Block (GCAB) and the Short Residual Block (SRB). The GCAB incorporates dilated convolutions to preserve overall features. It focuses on the important characteristics of each arrhythmia class from the original electrocardiogram data during the training process. The SRB employs a residual structure to enhance classification accuracy. The network’s performance is evaluated using a large-scale 12-lead electrocardiogram database for arrhythmia study on PhysioNet and the 2018 China Physiological Signal Challenge (CPSC) dataset. In particular, the proposed classification network shows the highest scores in average precision, recall, F1 score, area under the receiver operating characteristic, and accuracy compared to existing CNN-based arrhythmia classification networks in a large-scale 12-lead electrocardiogram database for arrhythmia study on PhysioNet. Finally, to evaluate the performance of the proposed classification network, we compared our proposed network with widely known classification networks such as VGGNet, ResNet, SENet, MobileNet, and EfficientNet. The proposed network demonstrates superior performance compared to other well-known classification networks. We validate the proposed arrhythmia classification network through confusion matrix and AUROC curve.
