A 2D post-beamforming filter for contrast restoration in medical ultrasound: in vivo results.
A 2D post-beamforming filter for contrast restoration in medical ultrasound: in vivo results.
复制标题
用于医学超声对比度恢复的 2D 后波束形成滤波器：体内结果。
DOI：
10.1109/iembs.2009.5333460
复制
发表时间：
2009
期刊：
影响因子：
--
通讯作者：
Ebbini,EmadS
中科院分区：
文献类型：
--
作者：
Wan,Yayun;Ebbini,EmadS
我们最近开发了一种强大的 2D 后波束形成滤波器，用于使用粗采样阵列孔径的超声成像系统中的对比度恢复，例如高频超声（HFUS）。该滤波器可以从感兴趣区域 (ROI) 中的离散 2D 脉冲响应模型导出。正则化二维伪逆滤波器鲁棒性的关键是将算子转换为 k 空间，其中正则化逆是使用 2D DFT 而不是计算上难以处理的矩阵运算来实现的。使用计算机模拟，2D PIO 被证明可以完全恢复由于 25-35 MHz 范围内 HFUS 阵列的粗略 2lambda 采样产生的栅瓣而导致的对比度损失。在本文中，我们首次展示了 2D PIO 使用市售探针对颈动脉进行成像的体内演示。结果表明，2D PIO 将组织/血液对比度提高了 4 dB（对血管横截面进行成像时）。这些结果与使用相同探针在成像质量保证体模中获得的实验结果一致。 2D PIO 消除栅瓣杂波的能力有望提高用于估计体内组织和血液位移的散斑跟踪算法的性能。
We have recently developed a robust 2D post-beamforming filter for contrast restoration in ultrasound imaging systems using coarsely-sampled array apertures, e.g. high frequency ultrasound (HFUS). The filter can be derived from a discretized 2D impulse response model in the region of interest (ROI). The key to the robustness of the regularized 2D pseudoinverse filter is transforming the operator to k-space, where the regularized inversion is implemented using 2D DFT instead of computationally intractable matrix operations. Using computer simulations, the 2D PIO was shown to produce complete restoration of contrast loss due to grating lobes resulting from coarse, 2lambda sampling of HFUS arrays in the 25-35 MHz range. In this paper, we present the first in vivo demonstration of the 2D PIO in imaging the carotid artery using a commercially available probe. The results show that the 2D PIO increases the tissue/blood contrast by 4 dB (when imaging a cross section of the vessel). These results are in agreement with experimental results obtained using the same probe in imaging quality assurance phantoms. The 2D PIO's ability to remove the clutter from grating lobe is expected to improve the performance of speckle tracking algorithms for the estimation of tissue and blood displacements in vivo.