心脏杂音的自适应时频谱分析

TIME-FREQUENCY ANALYSIS OF CARDIAC MURMURS USING ADAPTIVE CONE-KERNEL DISTRIBUTION

应用自适应时频分析方法对心脏杂音（收缩期杂音、舒张期杂音、连续性杂音）进行分类研究，旨在克服固定核时频分析分辨率较低的缺陷。对多例心脏病人不同类型的心脏杂音分析结果表明：基于自适应锥形核分布的时频谱反映了心脏杂音在时间－频率平面的能量分布及动态变化过程，并有较高的时频分辨率，不同类型心脏杂音的时频谱具有明显的时频特征。

The Cardiac murmurs (systolic murmurs, diastolic murmurs and mechanical murmurs) are classified by hemodynamics and studied by adaptive cone-kernel time-frequency analysis. The method can overcome the drawback of fix-kernel time-frequency analysis. The results, which are from the different kinds of murmurs in many patients with heart disease, show that the time-frequency spectrum based on adaptive cone-shaped kernel distribution are capable to correctly represent the power density and dynamic course of the series in time-frequency domain. The method has high time-frequency resolution, and the time-frequency spectrum of each murmur has its characteristics.