利用小波变换提取声谱图参数的一种方法

超声多普勒声谱技术在医学临床上应用时 ,往往需从声谱上计算声谱参数 ,这些参数已被用作医学诊断的依据。为了利用计算机对声谱图参数进行自动提取 ,本文提出了应用基于Mallat算法分析信号在小波变换后不同尺度性质的方法 ,在最大频率曲线上识别特征点 ,从而获得声谱图参数。应用该方法的仿真和临床实验都取得了较好的结果。

EXTRACTION OF SONOGRAM PARAMETERS USING WAVELET TRANSFORM

With the more and more clinical use of Doppler ultrasound technique, some sonogram parameters have been found to be related to blood flow situations. In order to extract sonogram parameters automatically, a multiscale analysis method based on Mallat algorithm was proposed. The recognition of characteristic points of a sonogram from its maximum frequency curve under different scales of the signal after the wavelet transform was described. The results of simulation and clinical use were both showed.