Enhancing dynamic ECG heartbeat classification with lightweight transformer model

作者：

Highlights：

• Taking on the challenging task of dynamic ECG heartbeat accurate classification.

• A novel light model, Lightweight Fussing Transformer, to address low data quality problem on dynamic ECG.

• An extensive experimental evaluation over shows that the techniques outperform previous approaches.

• Results indicate that the proposed methods are able to yield acceptable results and can be deployed in wearable device.

摘要

•Taking on the challenging task of dynamic ECG heartbeat accurate classification.•A novel light model, Lightweight Fussing Transformer, to address low data quality problem on dynamic ECG.•An extensive experimental evaluation over shows that the techniques outperform previous approaches.•Results indicate that the proposed methods are able to yield acceptable results and can be deployed in wearable device.

论文关键词：ECG classification,Arrhythmia detection,Attention,Transformer,Deep learning

论文评审过程：Received 22 June 2021, Revised 2 January 2022, Accepted 2 January 2022, Available online 7 January 2022, Version of Record 13 January 2022.

论文官网地址：https://doi.org/10.1016/j.artmed.2022.102236