融合特征空间最小方差波束形成和广义相干系数的超声成像方法

Eigenspace-based Minimum Variance Beamforming Combined with General Coherence Factor for Ultrasound Beamforming

为了进一步提高超声成像的质量,提出了融合特征空间最小波束形成和广义相干系数的成像方法。首先利用最小方差法计算回波数据的协方差矩阵和加权向量;然后对协方差矩阵进行特征分解得到信号子空间,并将加权向量投影到信号子空间,得到特征空间方法的加权向量;同时把阵元数据变换到波束域用于广义相干系数的计算,最后用广义相干系数作为加权系数对特征空间最小方差波束形成的结果进行优化。为了验证算法的有效性,对医学成像上常用的点目标和斑目标进行了成像,仿真实验结果表明:与特征空间最小方差算法和融合特征空间与相干系数的算法相比,本研究提出的方法提高了对比度以及稳健性,其代价是略微降低了成像分辨率。

To improve the quality of medical ultrasound imaging,a beamforming method which combines eigenspace-based mini-mum variance （ESBMV）with general coherence factor（GCF）was proposed.Firstly,minimum variance beamforming was used to obtain covariance matrix and weight vector;then the weight vector of the ESBMV was found by projecting the MV weight vector onto a vector subspace constructed from the eigenstructure of the covariance matrix;at the same time ,the data was transformed from array space to beamspace to calculate the general factor;in the end ,the general factor was used to optimize the results of eigenspace-based mini-mum variance beamforming.Simulations of point scatters and cyst phantom were used to verify the proposed method.The results show that the proposed method provides improved contrast,better speckle performance and more robustness than the ESBMV and ESBMV-CF beamforming method,at the expense of slightly lower resolution.