自适应波束形成在超声平面波造影的应用

Application of Adaptive Beamforming Technique in Plane-Wave Contrast-Enhanced Ultrasound Imaging

目前,超声平面波造影图像空间分辨率和信噪比的提升主要依赖于相干角度复合技术。然而,角度复合次数的增加,势必导致造影成像帧率的下降以及微泡击碎率的提升。为了克服该技术瓶颈,本文将具有更窄主瓣宽度和更强干扰/背景噪声抑制能力的自适应波束形成应用于超声平面波造影成像,力求以较少的角度复合次数达到与常规波束形成方法相当甚至更优的空间分辨率和信噪比。其中,采用前后向空域平滑方法对阵列协方差矩阵进行解相干处理,以抑制多次反射回波和相干杂波,最后引入对角加载技术以提升自适应方法的稳健性。本文使用迈瑞公司的Resona7平台配以L11-3U线阵探头,采集平面波造影通道数据进行MATLAB仿真,验证了所提出方法的有效性和性能优势。基于体模数据的仿真结果显示,在同为17次相干角度复合的条件下,本文方法的横向分辨率优于常规波束形成方法;基于犬肝脏活体数据(注射Sono Vue造影,剂量为0.05m L/kg)的仿真结果则表明,本文方法只需9次复合即可达到与常规方法13次复合相当的造影信噪比。

To improve spatial resolutions and signal-to-noise ratios（SNRs） of plane-wave contrast-enhanced ultrasound（PWCEUS） images, coherent angle compounding has been adequately investigated as an efficient way. Nevertheless, the increase of angle compounding number can inevitably result in a reduced frame rate and an elevated microbubble destruction. Hence in this paper, we applied the adaptive beamforming with a narrower main-beam width and enhanced ability of interference/noise suppression to PW-CEUS in order to obtain equivalent or even better performance as compared to the conventional beamforming algorithm with a smaller number of compounding angles. In particular, the forward-backward spatial smoothing was firstly applied to the covariance matrix for decoherence, and then the well-known diagonal loading technique was used to improve robustness of the adaptive beamformer. Simulation results had verified efficiency and superiority of the proposed algorithm by using the PW-CEUS channel data captured via Mindray Resona7 platform equipped with a linear probe L11-3 U. The phantom-based result indicates an improved lateral resolution of the proposed algorithm as compared to the traditional delay-and-sum（DAS） method, with a same number of compounding angles at 17. The simulation result based on dog-liver data demonstrates that, in order to yield an identical SNR, only 9 angles are needed for the proposed algorithm whereas 13 angles are needed for the traditional DAS method.