Department of Radiology, Thomas Jefferson University, 132 South 10th St., Philadelphia, PA 19107, USA.
Ultrason Imaging. 2012 Apr;34(2):81-92. doi: 10.1177/016173461203400202.
Subharmonic-aided pressure estimation (SHAPE) is a technique that utilizes the subharmonic emissions, occurring at half the insonation frequency, from ultrasound contrast agents to estimate ambient pressures. The purpose of this work was to compare the performance of different processing techniques for the raw radiofrequency (rf) data acquired for SHAPE. A closed loop flow system was implemented circulating reconstituted Sonazoid (GE Healthcare, Oslo, Norway; 0.2 ml for 750 ml diluent) and the beam-formed unprocessed rf data were obtained from a 4 mm diameter lumen of a Doppler flow phantom (ATS Laboratories, Inc., Bridgeport, CT) using a SonixRP scanner (Ultrasonix, Richmond, BC, Canada). The transmit frequency and incident acoustic pressures were set to 2.5 MHz and 0.22 MPa, respectively, in order to elicit Sonazoid subharmonic emissions that are ambient-pressure sensitive. The time-varying ambient pressures within the flow phantom were recorded by a Millar pressure catheter. Four techniques for extracting the subharmonic amplitude from the rf data were tested along with two noise filtering techniques to process this data. Five filter orders were tested for the noise removing filters. The performance was evaluated based on the least root-mean-square errors reported after linear least-square regression analyses of the subharmonic data and the pressure catheter data and compared using a repeated ANOVA. When the subharmonic amplitudes were extracted as the mean value within a 0.2 MHz bandwidth about 1.25 MHz and when the resulting temporally-varying subharmonic signal was median filtered with an order of 500, the filtered subharmonic signal significantly predicted the ambient pressures (r2 = 0.90; p < 0.001) with the least error. The resulting root mean square and mean absolute errors were 8.16 +/- 0.26 mmHg and 6.70 +/- 0.17 mmHg, respectively. Thus, median processing the subharmonic data extracted as the mean value within a 0.2 MHz bandwidth about the theoretical subharmonic frequency turned out to be the best technique to process acoustic data for SHAPE. The implementation of this technique on ultrasound scanners may permit real-time SHAPE applications.
亚谐波辅助压力估计 (SHAPE) 是一种利用超声对比剂在半倍超声入射频率下产生的亚谐波发射来估计环境压力的技术。本研究旨在比较不同处理技术对 SHAPE 采集的原始射频 (rf) 数据的性能。通过闭环流量系统循环重构成像剂（GE Healthcare，奥斯陆，挪威；750ml 稀释剂中 0.2ml），并使用 SonixRP 扫描仪（Ultrasonix，里士满，BC，加拿大）从多普勒流量体模（ATS Laboratories，Inc.，Bridgeport，CT）的 4mm 直径管腔中获取波束形成的未处理 rf 数据。发射频率和入射声压分别设置为 2.5MHz 和 0.22MPa，以产生对环境压力敏感的声诺维亚亚谐波发射。通过 Millar 压力导管记录流量体模内随时间变化的环境压力。测试了从 rf 数据中提取亚谐波幅度的四种技术，以及两种噪声滤波技术来处理此数据。对噪声消除滤波器测试了五个滤波器阶数。基于亚谐波数据和压力导管数据的线性最小二乘回归分析后报告的最小均方根误差来评估性能，并使用重复方差分析进行比较。当亚谐波幅度作为 1.25MHz 附近 0.2MHz 带宽内的平均值提取，并且随时间变化的亚谐波信号用 500 阶中位数滤波器滤波时，滤波后的亚谐波信号显著预测环境压力（r2=0.90;p<0.001），误差最小。所得均方根和平均绝对误差分别为 8.16+/-0.26mmHg 和 6.70+/-0.17mmHg。因此，将亚谐波数据作为 1.25MHz 附近 0.2MHz 带宽内的平均值提取并进行中位数处理的技术是处理 SHAPE 声数据的最佳技术。在超声扫描仪上实施该技术可能允许实时 SHAPE 应用。
