An effective data enhancement method for classification of ECG arrhythmia

Our blood vessels show signs of aging as we grow older, which leads to various cardiovascular diseases. Arrhythmia is usually the symptom of patients with early cardiovascular diseases. Early detection of arrhythmia is of great significance to the mortality of cardiovascular diseases. Applying deep learning to arrhythmia detection can help doctors discover cardiovascular diseases in time. At present, the performance of arrhythmia classification algorithms based on convolutional neural networks has far surpassed traditional methods. How-ever, the imbalance of arrhythmia data will seriously affect the performance of the classification algorithm. To better apply the convolutional neural network to the arrhythmia classification, a large amount of labeled ECG data is required. Therefore, this article proposes ECG Deep Convolution Generative Adversarial Networks (ECG-DCGAN) to expand the scarce data in the arrhythmia dataset and solve the problem of arrhythmia data imbal-ance. In addition, the convolution neural network (CNN) model is used to automatically classify the ECG signals without artificial feature extraction. Experimental results show that the classification method proposed in this paper improves the accuracy of arrhythmia diagnosis to 98.7% and that the algorithm used in this paper has good recognition performance and high clinical application value.

Automated summary

The study proposes ECG Deep Convolution Generative Adversarial Networks (ECG-DCGAN) to expand the arrhythmia dataset and solve the problem of data imbalance. Experimental results show that the proposed classification method significantly improves the accuracy of arrhythmia diagnosis.