Joint Department of Biomedical Engineering, University of North Carolina at Chapel Hill and North Carolina State University, Chapel Hill, NC 27599, USA.
Invest Radiol. 2010 Oct;45(10):669-74. doi: 10.1097/RLI.0b013e3181ef0a78.
Contrast-enhanced ultrasound imaging has demonstrated significant potential as a noninvasive technology for monitoring blood flow in the microvasculature. With the application of nondestructive contrast imaging pulse sequences combined with a clearance-refill approach, it is possible to create quantitative time-to-refill maps of tissue correlating to blood perfusion rate. One limitation to standard two-dimensional (2D) perfusion imaging is that the narrow elevational beamwidth of 1- or 1.5-D ultrasound transducers provides information in only a single slice of tissue, and thus it is difficult to image exactly the same plane from study to study. We hypothesize that inhomogeneity in vascularization, such as that common in many types of tumors, makes serial perfusion estimates inconsistent unless the same region can be imaged repeatedly. Our objective was to evaluate error in 2D quantitative perfusion estimation in an in vivo sample volume because of differences in transducer positioning. To mitigate observed errors due to imaging plane misalignment, we propose and demonstrate the application of quantitative 3-dimensional (3D) perfusion imaging. We also evaluate the effect of contrast agent concentration and infusion rate on perfusion estimates.
Contrast-enhanced destruction-reperfusion imaging was performed using parametric mapping of refill times and custom software for image alignment to compensate for tissue motion. Imaging was performed in rats using a Siemens Sequoia 512 imaging system with a 15L8 transducer. A custom 3D perfusion mapping system was designed by incorporating a computer-controlled positioning system to move the transducer in the elevational direction, and the Sequoia was interfaced to the motion system for timing of the destruction-reperfusion sequence and data acquisition. Perfusion estimates were acquired from rat kidneys as a function of imaging plane and in response to the vasoactive drug dopamine.
Our results indicate that perfusion estimates generated by 2D imaging in the rat kidney have mean standard deviations on the order of 10%, and as high as 22%, because of differences in initial transducer position. This difference was larger than changes in kidney perfusion induced by dopamine. With application of 3D perfusion mapping, repeatability in perfusion estimated in the kidney is reduced to a standard deviation of less than 3%, despite random initial transducer positioning. Varying contrast agent administration rate was also observed to bias measured perfusion time, especially at low concentrations; however, we observed that contrast administration rates between 2.7 × 10(8) and 3.9 × 10(8) bubbles/min provided results that were consistent within 3% for the contrast agent type evaluated.
Three-dimensional perfusion imaging allows a significant reduction in the error caused by transducer positioning, and significantly improves the reliability of quantitative perfusion time estimates in a rat kidney model. When performing perfusion imaging, it is important to use appropriate and consistent contrast agent infusion rates to avoid bias.
对比增强超声成像已显示出作为监测微血管血流的非侵入性技术的巨大潜力。通过应用无损对比成像脉冲序列和清除-再填充方法，可以创建与血流灌注率相关的组织的定量再填充时间图。标准二维（2D）灌注成像的一个限制是 1 或 1.5-D 超声换能器的窄轴向波束宽度仅在组织的单个切片中提供信息，因此很难从一项研究到另一项研究准确地对同一平面进行成像。我们假设，血管化的不均匀性，如许多类型的肿瘤中常见的那样，除非可以重复成像相同的区域，否则会导致连续的灌注估计不一致。我们的目标是评估由于换能器定位不同而导致的体内样本体积中的 2D 定量灌注估计的误差。为了减轻由于成像平面未对准而引起的观察到的误差，我们提出并演示了定量 3 维（3D）灌注成像的应用。我们还评估了对比剂浓度和输注率对灌注估计的影响。
使用再填充时间的参数映射和用于图像对准的定制软件进行对比增强破坏-再灌注成像，以补偿组织运动。使用西门子 Sequoia 512 成像系统和 15L8 换能器在大鼠中进行成像。通过结合计算机控制的定位系统来在垂直方向上移动换能器，设计了定制的 3D 灌注映射系统，并将 Sequoia 与运动系统接口连接，以用于破坏-再灌注序列的定时和数据采集。根据成像平面并响应血管活性药物多巴胺，从大鼠肾脏中获取灌注估计值。
我们的结果表明，由于初始换能器位置的差异，大鼠肾脏中 2D 成像生成的灌注估计值的平均值标准偏差约为 10%，最高可达 22%。这种差异大于多巴胺诱导的肾脏灌注变化。应用 3D 灌注映射后，尽管初始换能器位置随机，但在肾脏中估计的灌注的可重复性降低至小于 3%的标准偏差。还观察到，改变对比剂给药率会使测量的灌注时间产生偏差，尤其是在低浓度下；但是，我们观察到，在所评估的对比剂类型下，以 2.7×10^8 和 3.9×10^8 个气泡/分钟之间的对比剂给药率提供的结果在 3%以内是一致的。
3D 灌注成像可以显著降低换能器定位引起的误差，并显著提高大鼠肾脏模型中定量灌注时间估计的可靠性。进行灌注成像时，使用适当且一致的对比剂输注率非常重要，以避免偏差。
