A novel hybrid energy fraction and entropy-based approach for systolic heart murmurs identification
作者：
Highlights：
• We proposed a novel feature extraction based on wavelet packet and sample entropy.
• We proposed five novel feature descriptions of heart sounds.
• We discovered the relationship between energy fraction and area of murmurs.
• Optimal kernel function and parameters for murmurs classification were selected.
• This paper reviewed and compared recently reports on heart sound classification.
摘要
•We proposed a novel feature extraction based on wavelet packet and sample entropy.•We proposed five novel feature descriptions of heart sounds.•We discovered the relationship between energy fraction and area of murmurs.•Optimal kernel function and parameters for murmurs classification were selected.•This paper reviewed and compared recently reports on heart sound classification.
论文关键词：Systole murmurs classification,WP decomposition,Selective reconstitution,Energy fraction,S1–S2sampen and HMsampen,SVM
论文评审过程：Available online 8 November 2014.
论文官网地址：https://doi.org/10.1016/j.eswa.2014.10.051
