Extracting deep features from short ECG signals for early atrial fibrillation detection
作者：
Highlights：
• Atrial Fibrillation at early stage has no obvious symptoms and is difficult to detect.
• Extracting effective features is critical to AF detection using machine learning approaches.
• Existing studies used shallow time, frequency or energy features with weak representation.
• We propose three deep features that can accurately capture the subtle variation in short ECG segment.
• With the proposed feature set we achieve outstanding classification results than others.
摘要
•Atrial Fibrillation at early stage has no obvious symptoms and is difficult to detect.•Extracting effective features is critical to AF detection using machine learning approaches.•Existing studies used shallow time, frequency or energy features with weak representation.•We propose three deep features that can accurately capture the subtle variation in short ECG segment.•With the proposed feature set we achieve outstanding classification results than others.
论文关键词：Medical knowledge engineering,Deep features extraction,Early atrial fibrillation detection,Data mining
论文评审过程：Received 11 November 2019, Revised 18 May 2020, Accepted 29 May 2020, Available online 3 June 2020, Version of Record 22 September 2020.
论文官网地址：https://doi.org/10.1016/j.artmed.2020.101896
