基于微泡母小波变换的超声造影成像方法

Ultrasound Contrast Agent Imaging Based on Bubble Wavelet Transform

目的利用微泡母小波技术(bubble wavelet transform,BWT)提高平面波超声造影成像质量。方法首先基于Doinikov微泡模型构建新型微泡母小波,其次将BWT技术拓展到频域,根据尺度因子所对应的中心频率找到最优尺度,避免了目前需要遍历选择尺度因子所产生的计算量。最后在前端BWT技术的基础上对比现有后端波束形成算法的成像结果。结果仿体实验结果说明,当尺度因子选在其对应的中心频率为微泡二次谐波处时所得到的组织比(contrast-to-tissue ratio,CTR)最高。最优尺度下使用BWT前后,3 MHz发射频率下延迟叠加算法(delay-and-sum,DAS),最小方差算法(minimum variance,MV)及特征空间最小方差算法(eigenspace based minimum variance,ESBMV)的CTR分别提高了1.05d B,1.04 d B和22.02 d B;4 MHz频率下分别提高了0.9 d B,1.04 d B和18.84 d B。结论 BWT技术中最优尺度理应选在其对应的中心频率为微泡的二次谐波处。接收端波束形成方面,相对DAS和MV算法,BWT技术结合ESBMV算法得到的成像质量可以进一步提高。

Objective To improve the quality of ultrasound plane wave contrast-enhanced imaging based on bubble wavelet transform（ BWT）. Methods A new bubble wavelet was firstly constructed according to Doinikov model. Then we gave the theoretical basis for the choice of the optimal scale according to the corresponding central frequency and avoided the amount of computation. Last,we compared the imaging results of the existing receiving-end beamforming algorithms on the basis of BWT technology. Results The phantom experiment results suggested that the highest contrast-to-tissue ratio（ CTR） value was obtained when the scale factor was selected with the center frequency falling at the second harmonic of microbubble. The image CTR of delayand-sum（ DAS）,minimum variance（ MV） and eigenspace based minimum variance（ ESBMV） beamforming improved 1. 05 d B,1. 04 d B and 22. 02 d B under 3 MHz while 0. 9 d B,1. 04 d B and 18. 84 d B under 4 MHz respectively. Conclusion The optimal scale factor should be selected with the corresponding center frequency falling at the second harmonic of microbubble. For the receiving-end beamforming,a better image quality can be obtained by combining BWT and ESBMV beamformer compared with that of DAS and MV.