浅表组织超声图像的均衡化处理

Equalization of Superficial Ultrasonic Images

均衡化算法可有效降低超声图像采样点间的相关性,提高斑点噪声的抑制效果。但根据在体图像获取均衡滤波器参数的计算量大,不利于实时实现。针对浅表超声图像,本研究提出一种效果明显同时避免大量实时计算的均衡化方法。基于射频图像成像模型,分析超声体模回波信号频谱,滤除组织反射函数频谱成分,获得点扩散函数(PSF),构建均衡滤波器,分析其用于浅表组织超声图像的均衡化处理效果。通过超声体模获取参数的均衡滤波器用于浅表超声图像,可以有效降低采样点间的相关性,预处理后的去斑滤波效果相较于直接滤波,信噪比提高约19.2%,图像去斑效果得到提高。对于浅表超声图像,可以通过超声体模预先获取均衡滤波器参数,从而减少去斑过程的计算量,容易实时实现。

Equalization algorithm can effectively reduce the correlation between ultrasonic image samples; improve the performance of de-speckling methods. But acquiring equalization filter parameters based on in vivo images requires a large amount of calculation which is not conducive to real-time processing. Aiming at superficial ultrasonic images, this paper gives an effective equalization method that avoids lots of real-time calculation. Based on the ultrasonic RF imaging formation model, this paper analyzed the spectrum of ultrasonic phantom echo signals, filtered the spectral component of the tissue reflectivity function, obtained the point- spread function and constructed the equalization filter to analyze its effect for superficial images. With the parameters acquired from the phantom the equalization filter used for superficial ultrasonic images can reduce the correlation. Compared with direct de-speckling methods, the speckle SNR of preprocessed de-speckling methods has been improved about 19.2%, the performance of de-speckling methods gets improved. For superficial ultrasonic images, parameters of equalization filter can be acquired through ultrasound phantom in advance, it can reduce the calculation of de-speckling process and easy to realize real-time processing.