Fully automatic electrocardiogram classification system based on generative adversarial network with auxiliary classifier
作者：
Highlights：
• Proposed a fully automatic generative adversarial network based system for arrhythmia screening.
• High F1 score for supraventricular ectopic beat detection against the state-of-arts.
• Our automatic method performs at comparative level as expert-assisted methods.
• Adopting discriminator as classifier after fine-tuning boosts final performance.
• Data augmentation helps relieve class imbalance problem of arrhythmia detection.
摘要
•Proposed a fully automatic generative adversarial network based system for arrhythmia screening.•High F1 score for supraventricular ectopic beat detection against the state-of-arts.•Our automatic method performs at comparative level as expert-assisted methods.•Adopting discriminator as classifier after fine-tuning boosts final performance.•Data augmentation helps relieve class imbalance problem of arrhythmia detection.
论文关键词：Generative adversarial network,Arrhythmia,ECG classification,Data augmentation
论文评审过程：Received 19 June 2020, Revised 21 December 2020, Accepted 27 February 2021, Available online 3 March 2021, Version of Record 13 March 2021.
论文官网地址：https://doi.org/10.1016/j.eswa.2021.114809
