Searching for Premature Ventricular Contraction from Electrocardiogram by Using One-Dimensional Convolutional Neural Network

Abstract

Premature ventricular contraction (PVC) is a common cardiac arrhythmia that can occur in ordinary healthy people and various heart disease patients. Clinically, cardiologists usually use a long-term electrocardiogram (ECG) as a medium to detect PVC. However, it is time-consuming and labor-intensive for cardiologists to analyze the long-term ECG accurately. To this end, this paper suggests a simple but effective approach to search for PVC from the long-term ECG. The recommended method first extracts each heartbeat from the long-term ECG by applying a fixed time window. Subsequently, the model based on the one-dimensional convolutional neural network (CNN) tags these heartbeats without any preprocessing, such as denoise. Unlike previous PVC detection methods that use hand-crafted features, the proposed plan rationally and automatically extracts features and identify PVC with supervised learning. The proposed PVC detection algorithm acquires 99.64% accuracy, 96.97% sensitivity, and 99.84% specificity for the MIT-BIH arrhythmia database. Besides, when the number of samples in the training set is 3.3 times that of the test set, the proposed method does not misjudge any heartbeat from the test set. The simulation results show that it is reliable to use one-dimensional CNN for PVC recognition. More importantly, the overall system does not rely on complex and cumbersome preprocessing.

Premature Ventricular Contraction Detection Based on Chebyshev Polynomials and K Nearest Neighbours Classifier

“…On the other hand, the proposed approach was not superior to the references [16,24,34] in terms of specificity and accuracy. Second, the accuracy of the suggested approach is found to be good, with only 0.34% less than the reference [24].…”

“…ECG classification is a difficult task due to the significant variations in ECG signals for different patients. In recent years, several algorithms for ECG classification, heartbeat detection, and diagnosis have been proposed [13][14][15][16]. The PhysioNet/CINC 2020 and 2021 Challenges [17,18] provide an opportunity to discuss the complexities of ECG classification from several perspectives and the impact of analysing large numbers of leads.…”

“…The PhysioNet/CINC 2020 and 2021 Challenges [17,18] provide an opportunity to discuss the complexities of ECG classification from several perspectives and the impact of analysing large numbers of leads. Algorithms for ECG classification can be divided into two groups: morphologybased methods [1,[13][14][15][19][20][21][22][23] and deep learning-based methods [16,[24][25][26][27][28][29][30].…”

“…Conversely, the use of deep learning has been prevalent in identifying PVCs in ECG signals and several investigations have demonstrated various deep learning approaches. An example of this is the use of one-dimensional convolutional neural networks (CNNs) for feature extraction and the identification of PVCs via supervised learning [16]. Yu et al [24] used deep metric learning and a k-nearest neighbours (KNN) classifier to detect PVCs in long-term ECG signals, while Zhou et al [25] employed a recurrent neural network featuring long short-term memory to identify PVCs in ECG signals.…”

An Automatic Premature Ventricular Contraction Recognition System Based on Imbalanced Dataset and Pre-Trained Residual Network Using Transfer Learning on ECG Signal

“…With the growing interest of biometric authentication, deep learning plays an important role as a reliable tool to the function easily and accurately from ECGs [32]. Searching of PVCs from the time-series ECG signals with 1D CNNs is a common detection approach and provides good findings in some scenarios [33]. Herein, features are extracted automatically compared to hand-crafted feature extraction methods discussed aforementioned and a Softmax classifier is usually used for the classification in the last layer.…”