Why the study?
Existing arrhythmia classification methods still fail to provide acceptable performance when detecting different types of cardiac abnormalities, particularly with unbalanced datasets.
Does a deep WaveNet-LSTM convolution model improve the accuracy of ECG heartbeat classification for arrhythmia detection?
Does a deep WaveNet-LSTM convolution model improve the accuracy of ECG heartbeat classification for arrhythmia detection?
A novel WaveNet-LSTM convolution model achieves 96.8% accuracy in detecting abnormal ECG signals for arrhythmia classification.
Electrocardiogram (ECG) can effectively record the potential difference of the body surface generated during the physiological function of the heart, and analyze and accurately discriminate the potential difference signal generated by the electrocardiogram, which can effectively prevent sudden diseases and reduce suddenness. ECG heartbeat is a powerful tool to diagnose several abnormal arrhythmias. Although the classification methods of arrhythmia improved significantly, but when detecting different types of cardiac abnormalities, especially when dealing with unbalanced data sets, it still does not provide acceptable performance. In this paper, we propose a method of data processing combined with the deep WaveNet-LSTM convolution model to solve this limitation of the current classification method. Furthermore, to prove the effectiveness of proposed method, using the MIT-BIH arrhythmia database, which considers intra- and inter-patient paradigms, and the AAMI EC57 standard. The evaluation results show that our method has achieved very good performance, and its accuracy of ECG abnormal signal detection reaches 96.8%.
No takes yet. Share an insight, caveat, or question.
Qu et al. (2020) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: