Atrial Fibrillation Detection Using an Improved Multi-Scale Decomposition Enhanced Residual Convolutional Neural Network

Xin-Cheng Cao,Bin-Qiang Chen,Bin Yao

doi:10.1109/access.2019.2926749

Xin-Cheng Cao, Bin-Qiang Chen + Show 1 more

Open Access

https://doi.org/10.1109/access.2019.2926749

Copy DOI

Abstract

Atrial fibrillation, the most common sustained arrhythmia, is still a big challenge for researchers in the medical field. Many studies attempt to realize intelligent classification of AF based on deep learning methods. However, many of the studies focused on investigations of relatively simple datasets collected from a relatively small number of subjects. On the other hand, sophisticated preprocessing is usually adopted to analyse the ECG signals. These two factors significantly affect the generalization ability of the trained models for complicated data sets collected from a large number of subjects. In order to address this problem, an improved multi-scale decomposition enhanced residual convolutional neural network is proposed. The proposed method is applied to the large single-lead ECG dataset provided by the PhysioNet/CinC Challenge 2017, and good classification accuracy is suggested by the testing results. In the proposed method, the original ECG record with a large difference in length is re-segmented into short samples of 9 s. Then, using the derived wavelet frame decomposition, the segmented short samples are decomposed and reconstituted into sub-signal samples of different scales. We trained the fast down-sampling residual convolutional neural networks (FDResNets) with the original short-signal dataset and the reconstructed dataset of each scale. The transfer learning technique is then applied to couple the three FDResNets with good performance into a multi-scale decomposition enhanced residual convolutional neural network (MSResNet). The FDResNet trained by the [0, 9.375 Hz] reconstruction dataset achieved the best performance. After six-fold cross-validation, the average test accuracy reached 87.12%, and the average comprehensive F1 score reached 85.29%. The average test accuracy of the multi-scale residual neural network reached 92.1%, and the average overall F1 score reached 89.9%.

Highlights

Atrial fibrillation (AF), the most common sustained arrhythmia, represents a difficult scientific challenge and remains enigmatic even after more than one century of research [1]
The results of the six-fold cross-validation of the FDResNets using down-sampling modules containing different number of wide-stride convolution (WSConv) layer are shown in Figure 6, and the number of epoch is 75
THE EFFECTIVENESS OF THE PROPOSED METHOD This paper aims at classifying the short single lead electrocardiogram arrhythmia, especially to identify whether the subject has atrial fibrillation based on their ECG recording

Summary

Introduction

Atrial fibrillation (AF), the most common sustained arrhythmia, represents a difficult scientific challenge and remains enigmatic even after more than one century of research [1]. The mechanisms responsible for AF are still not fully understood and the treatment is very complicated. The symptoms of AF are abnormal contractions of the upper atrium; on the electrocardiogram, there is a sinus P wave loss. ECG analysis has become an important means of AF diagnosis. An automated analysis and classification system for ECG records can provide physicians with diagnostic recommendations or help patients monitor their own health status, which is important for improving medical efficiency and reducing medical costs. The traditional ECG classification methodology includes three steps, i.e., signal preprocessing, feature extraction and classification. The first step aims at eliminating various types of noises, including artifacts and baseline drift in the signal

Objectives

Results

Discussion

Conclusion