IEEE Trans Ultrason Ferroelectr Freq Control. 2017 Aug;64(8):1212-1222. doi: 10.1109/TUFFC.2017.2717792. Epub 2017 Jun 20.
Current methods of cardiac strain imaging at high frame rate suffer from motion matching artifacts or poor lateral resolution. Coherent compounding has been shown to improve echocardiographic image quality while maintaining a high frame rate, but has never been used to image cardiac strain. However, myocardial velocity can have an impact on coherent compounding due to displacements between frames. The objective of this paper was to investigate the feasibility and performance of coherent compounding for cardiac strain imaging at a low and a high myocardial velocity. Left-ventricular contraction in short-axis view was modeled as an annulus with radial thickening and circumferential rotation. Simulated radio-frequency channel data with a cardiac phased array were obtained using three different beamforming methods: single diverging wave, coherent compounding of diverging waves, and conventional focusing. Axial and lateral displacements and strains as well as radial strains were estimated and compared to their true value. In vivo feasibility of cardiac strain imaging with coherent compounding was performed and compared to single diverging wave imaging. At low myocardial velocities, the axial, lateral, and radial strain relative error for nine compounded waves (16.3%, 40.4%, and 18.9%) were significantly lower than those obtained with single diverging wave imaging (19.9%, 80.3%, and 30.6%) and closer to that obtained with conventional focusing (16.7%, 43.7%, and 16%). In vivo left-ventricular radial strains exhibited higher quality with nine compounded waves than with single diverging wave imaging. These results indicate that cardiac strain can be imaged using coherent compounding of diverging waves with a better performance than with single diverging wave imaging while maintaining a high frame rate, and therefore, has the potential to improve diagnosis of myocardial strain-based cardiac diseases.
目前,高帧率心脏应变成像方法存在运动匹配伪影或横向分辨率差的问题。相干复合已被证明可以在保持高帧率的同时提高超声心动图图像质量,但从未用于心脏应变成像。然而,由于帧间的位移,心肌速度会对相干复合产生影响。本文的目的是研究在低和高心肌速度下,相干复合用于心脏应变成像的可行性和性能。短轴视图中的左心室收缩被建模为一个具有径向增厚和圆周旋转的环形。使用三种不同的波束形成方法:单发散波、发散波的相干复合和传统聚焦,获得具有心脏相控阵的模拟射频通道数据。估计并比较了轴向和横向位移及应变以及径向应变与其真实值。进行了相干复合心脏应变成像的体内可行性研究,并与单发散波成像进行了比较。在低心肌速度下,九个复合波的轴向、横向和径向应变相对误差(16.3%、40.4%和 18.9%)明显低于单发散波成像(19.9%、80.3%和 30.6%),更接近传统聚焦(16.7%、43.7%和 16%)。体内左心室径向应变的九个复合波优于单发散波成像。这些结果表明,与单发散波成像相比,使用发散波的相干复合可以更好地进行心脏应变成像,同时保持高帧率,因此有可能改善基于心肌应变的心脏病的诊断。