An effective feature extraction method based on GDS for atrial fibrillation detection

• Published in: Journal of biomedical informatics Vol. 119; p. 103819
• Main Authors: Wang, Haiyan, Dai, Honghua, Zhou, Yanjie, Zhou, Bing, Lu, Peng, Zhang, Hongpo, Wang, Zongmin
• Format: Journal Article
• Language: English
• Published: Elsevier Inc 01.07.2021
• Subjects: Atrial fibrillation; DNN; Feature extraction; Gradient set; Information quantity features; Statistical distribution features; Information quantity features; Feature extraction; DNN; Atrial fibrillation; Gradient set; Statistical distribution features
• ISSN: 1532-0464; 1532-0480; 1532-0480
• DOI: 10.1016/j.jbi.2021.103819

[Display omitted] •An effective extraction method is proposed for atrial fibrillation detection in short-term ECG.•Obtain the gradient set (GDS) of equal-length ECG segments.•Extracting the statistical and information quantity features of the GDS.•No beat detection, no denoising and normalization, suitable for most classifiers.•A simple fully connected Deep Neural Network (DNN). Atrial fibrillation (AF) is a common and extremely harmful arrhythmia disease. Automatic detection of AF based on ECG helps accurate and timely detection of the condition. However, the existing AF detection methods are mostly based on complex signal transformation or precise waveform localization. This is a big challenge for complex, variable, and susceptible ECG signals. Therefore, we propose a simple feature extraction method based on gradient set (GDS) for AF detection. The method first calculates the GDS of the ECG segment and then calculates the statistical distribution feature and the information quantity feature of the GDS as the input of the classifier. Experiments on four databases include 146 subjects show that the feature extraction method for detecting AF proposed in this paper has the characteristics of simple calculation, noise tolerance, and high adaptability to all kinds of classifiers, and got the best performance on the DNN classifier we designed. Therefore, it is a good choice for feature extraction in AF detection.