IEEE Trans Ultrason Ferroelectr Freq Control. 2018 Mar;65(3):300-315. doi: 10.1109/TUFFC.2017.2682180. Epub 2017 Mar 14.
According to the linear acoustic theory, the channel data of a plane wave emitted by a linear array is a linear combination of the full data set of synthetic transmit aperture (STA). Combining this relationship with compressed sensing (CS), a novel CS based ultrasound beamforming strategy, named compressed sensing based synthetic transmit aperture (CS-STA), was previously proposed to increase the frame rate of ultrasound imaging without sacrificing the image quality for a linear array. In this paper, assuming linear transfer function of a pulse-echo ultrasound system, we derived and applied the theory of CS-STA for a slightly curved array and validated CS-STA in a convex array configuration. Computer simulations demonstrated that, in the convex array configuration, the normalized root-mean-square error between the beamformed radio-frequency data of CS-STA and STA was smaller than 1% while CS-STA achieved four-fold higher frame rate than STA. In addition, CS-STA was capable of achieving good image quality at depths over 100 mm. It was validated in phantom experiments by comparing CS-STA with STA, multielement synthetic transmit aperture (ME-STA), and the conventional focused method (focal depth = 110 mm). The experimental results showed that STA and CS-STA performed better than ME-STA and the focused method at small depths. At the depth of 110 mm, CS-STA, ME-STA, and the focused methods improved the contrast and contrast-to-noise ratio of STA. The improvements in CS-STA are higher than those in ME-STA but lower than those in the focused mode. These results can also be observed qualitatively in the in vivo experiments on the liver of a healthy male volunteer. The CS-STA method is thus proved to increase the frame rate and achieve high image quality at full depth in the convex array configuration.
根据线性声学理论，线性阵列发射的平面波的通道数据是合成发射孔径（STA）全数据集的线性组合。将这种关系与压缩感知（CS）相结合，先前提出了一种基于 CS 的超声波束形成策略，称为基于压缩感知的合成发射孔径（CS-STA），用于在不牺牲线性阵列图像质量的情况下提高超声成像的帧率。在本文中，假设脉冲回波超声系统的线性传递函数，我们推导出并应用了 CS-STA 的理论用于略微弯曲的阵列，并在凸阵配置中验证了 CS-STA。计算机模拟表明，在凸阵配置中，CS-STA 和 STA 的波束形成射频数据之间的归一化均方根误差小于 1%，而 CS-STA 的帧率比 STA 高四倍。此外，CS-STA 能够在超过 100mm 的深度实现良好的图像质量。通过将 CS-STA 与 STA、多元合成发射孔径（ME-STA）和传统聚焦方法（聚焦深度=110mm）进行比较，在体模实验中验证了 CS-STA。实验结果表明，STA 和 CS-STA 在小深度处的性能优于 ME-STA 和聚焦方法。在 110mm 深度处，CS-STA、ME-STA 和聚焦方法提高了 STA 的对比度和对比噪声比。CS-STA 的改进高于 ME-STA，但低于聚焦模式。在健康男性志愿者肝脏的体内实验中也可以定性地观察到这些结果。因此，CS-STA 方法被证明可以在凸阵配置中提高帧率并在全深度实现高质量图像。
