Du Juan, Liu Dalong, Ebbini Emad S
IEEE Trans Ultrason Ferroelectr Freq Control. 2016 Dec;63(12):2069-2081. doi: 10.1109/TUFFC.2016.2614430. Epub 2016 Sep 29.
A nonlinear filtering approach to imaging the dynamics of microbubble ultrasound contrast agents (UCAs) in microvessels is presented. The approach is based on the adaptive third-order Volterra filter (TVF), which separates the linear, quadratic, and cubic components from beamformed pulse-echo ultrasound data. The TVF captures polynomial nonlinearities utilizing the full spectral components of the echo data and not from prespecified bands, e.g., second or third harmonics. This allows for imaging using broadband pulse transmission to preserve the axial resolution and the SNR. In this paper, we present the results from imaging the UCA activity in a 200- [Formula: see text] cellulose tube embedded in a tissue-mimicking phantom using a linear array diagnostic probe. The contrast enhancement was quantified by computing the contrast-to-tissue ratio (CTR) for the different imaging components, i.e., B-mode, pulse inversion (PI), and the TVF components. The temporal mean and standard deviation of the CTR values were computed for all frames in a given data set. Quadratic and cubic images, referred to as QB-mode and CB-mode, produced higher mean CTR values than B-mode, which showed improved sensitivity. Compared with PI, they produced similar or higher mean CTR values with greater spatial specificity. We also report in vivo results from imaging UCA activity in an implanted LNCaP tumor with heterogeneous perfusion. The temporal means and standard deviations of the echogenicity were evaluated in small regions with different perfusion levels in the presence and absence of UCA. The in vivo measurements behaved consistently with the corresponding calculations obtained under microflow conditions in vitro. Specifically, the nonlinear VF components produced larger increases in the temporal mean and standard deviation values compared with B-mode in regions with low to relatively high perfusion. These results showed that polynomial filters such as the TVF can provide an important tool for imaging UCA activity in regions with heterogeneous perfusion as is the case in some tumors and ischemic tissues.
提出了一种用于成像微血管中微泡超声造影剂(UCA)动力学的非线性滤波方法。该方法基于自适应三阶沃尔泰拉滤波器(TVF),它从波束形成的脉冲回波超声数据中分离出线性、二次和三次分量。TVF利用回波数据的全频谱分量而非预先指定的频段(如二次或三次谐波)来捕捉多项式非线性。这使得能够使用宽带脉冲传输进行成像,以保持轴向分辨率和信噪比。在本文中,我们展示了使用线性阵列诊断探头对嵌入组织模拟体模中的200 - [公式:见正文]纤维素管内的UCA活性进行成像的结果。通过计算不同成像分量(即B模式、脉冲反转(PI)和TVF分量)的对比组织比(CTR)来量化对比增强。计算给定数据集中所有帧的CTR值的时间平均值和标准差。二次和三次图像(称为QB模式和CB模式)产生的平均CTR值高于B模式,显示出更高的灵敏度。与PI相比,它们产生的平均CTR值相似或更高,且具有更高的空间特异性。我们还报告了对植入的具有异质灌注的LNCaP肿瘤中UCA活性进行成像的体内结果。在存在和不存在UCA的情况下,在具有不同灌注水平小区域中评估回声性的时间平均值和标准差。体内测量结果与在体外微流条件下获得的相应计算结果一致。具体而言,在低灌注到相对高灌注区域,与B模式相比,非线性VF分量在时间平均值和标准差上产生更大的增加。这些结果表明,诸如TVF之类的多项式滤波器可以为成像某些肿瘤和缺血组织中存在的异质灌注区域的UCA活性提供重要工具。