Automatic diagnosis of arrhythmia with electrocardiogram using multiple instance learning: From rhythm annotation to heartbeat prediction

作者：

Highlights：

• Develop an arrhythmia classifier for heartbeats using ECG rhythm annotations

• Develop a two-stage algorithm for multiple instance learning (MIL)

• Validate the algorithms using a large real-world ECG dataset

摘要

•Develop an arrhythmia classifier for heartbeats using ECG rhythm annotations•Develop a two-stage algorithm for multiple instance learning (MIL)•Validate the algorithms using a large real-world ECG dataset

论文关键词：Electrocardiogram,Multiple instance learning,Deep learning,Heartbeat classification

论文评审过程：Received 12 November 2021, Revised 30 June 2022, Accepted 18 August 2022, Available online 22 August 2022, Version of Record 30 August 2022.

论文官网地址：https://doi.org/10.1016/j.artmed.2022.102379