Multi-domain Feature Fusion Neural Network for Electrocardiogram Classification.

Computerized electrocardiogram (ECG) interpretation technology is of great significant in detecting cardiovascular disease. Recently, many research studies information mining of ECG signals from multiple domain. However, these models of automatic arrhythmia detection extract the features in time-domain and time-frequency domain separately, and merge the features in the last layer. The information of middle layers of multi-domain is not used fully. In this study, we develop an ECG classifier based on Multiple Domain Features Fusion Network with Lead Attention (MDFF-LA), which can realize feature fusion in time domain and frequency domain features respectively. Filter useful lead information through the attention module prior to data entry. In the process of feature extraction, the feature weights of one domain are optimized in multiple stages by the attention module generated by another domain, where the multi-domain information constantly complements each other at the middle layers. Finally, we conduct comprehensive experiments on three multi lead ECG databases to test the performance. The results demonstrate that the fusion features in multi domain extracted by MDFF-LA can obtain more valuable information, which can provide supported diagnosis for clinicians in practical.