Ecg heartbeat classification based on resnet and bi-lstm

Abstract

As a reliable cardiovascular system, ECG has been widely used in the detection of heart rhythms. This paper presents an attention-based ResNet processing of ECG data. Respectively are MIT and PTB Diagnostics datasets. The model is prominent in both data sets. And the accuracy in MIT datasets is 96.2%. And in PTB datasets the accuracy is 99.6%.