51黑料吃瓜在线观看,51黑料官网|51黑料捷克街头搭讪_51黑料入口最新视频

Objective Heart murmur of heart sounds can reflect the abnormal defects of the heart structure. A heart sound segmentation and analysis method is proposed for heart murmur identification of congenital heart disease. Methods At first, a novel noise reduction method based on frequency slice wavelet transform (FSWT) is applied to remove the surrounding noise. And then the duration of heart murmur is segmented and extracted by the characteristic moment waveform. Finally, the Mel-scale frequency features are extracted based on Mel frequency cepstrum coefficients for the heart murmur signal. Results The heart sound data including 32 normal cases and 16 ventricular septal defect (VSD) cases are applied to validate the efficiency of the proposed method. And the parameter scopes of normal and VSD data are calculated by the average values and standard deviations of the Mel-scale features. Conclusions The proposed method can identify the characteristic of VSD heart sounds in the duration of systolic murmurs. And more features may be extracted from diastolic murmurs for further analysis. The proposed method can be the effective reference for the screening and monitoring of typical congenital heart disease.