Stadler R W, Taylor J A, Lees R S
Boston Heart Foundation, Cambridge, MA 02142, USA.
Ultrasound Med Biol. 1997;23(6):879-87. doi: 10.1016/s0301-5629(97)00074-4.
Measurements of arterial diameter throughout the cardiac cycle (i.e., the arterial distension waveform) are conducted increasingly to study mechanical properties of the arterial wall and changes associated with disease. The distension waveform of peripheral arteries can be measured noninvasively via ultrasonic echo tracking. M-mode imaging, and B-mode imaging. Of these, echo tracking is the most popular method because of its single micrometer resolution during continuous measurements under ideal conditions. However, high resolution within continuous measurements does not imply high reproducibility between measurements. Therefore, we compared repeated measurements of the amplitude of common carotid artery distension in 26 subjects, obtained sequentially in random order by: 1. Off-line echo tracking of digitized radiofrequency ultrasound; 2. M-mode imaging with automated edge detection; and 3. 30-Hz B-mode imaging with automated edge detection and model-based diameter estimation. In each case, the transducer was hand-held and was removed from the neck between repeated measurements. The amplitude of arterial distension was estimated from the serial diameter measurements by maximum likelihood (ML) estimation, by least-squares fit of a Fourier series model, and by application of a cubic smoothing spline. Within continuous measurements, the standard deviation of the ML distension amplitude for neighboring cardiac cycles was significantly smaller (p > 0.05) with echo-tracking (0.023 mm) than with the B-mode (0.036 mm) or M-mode (0.074 mm) methods. However, between discontinuous measurements on the same subject, the standard deviation of the ML distension amplitude was similar for the echo-tracking (0.076 mm) and B-mode (0.073 mm) methods. The Fourier series model and the cubic smoothing spline slightly reduced the standard deviation of the B-mode and M-mode distension amplitudes, but also reduced the mean amplitude estimate. On the basis of this relative comparison of methods, we conclude that, although echo tracking offers high resolution for continuous measurements, the reproducibility of discontinuous measurements of carotid artery distension is no better with echo tracking than can be obtained from 30-Hz B-mode images.
为了研究动脉壁的力学特性以及与疾病相关的变化，越来越多地开展整个心动周期中动脉直径的测量（即动脉扩张波形测量）。外周动脉的扩张波形可通过超声回波跟踪、M 型成像和 B 型成像进行无创测量。其中，回波跟踪是最常用的方法，因为在理想条件下连续测量时它具有单微米分辨率。然而，连续测量中的高分辨率并不意味着测量之间具有高重现性。因此，我们比较了 26 名受试者颈总动脉扩张幅度的重复测量结果，这些测量结果是通过以下方式按随机顺序依次获得的：1. 对数字化射频超声进行离线回波跟踪；2. 采用自动边缘检测的 M 型成像；3. 采用自动边缘检测和基于模型的直径估计的 30Hz B 型成像。在每种情况下，换能器均为手持，且在重复测量之间从颈部移开。通过最大似然（ML）估计、傅里叶级数模型的最小二乘拟合以及应用三次平滑样条，从连续的直径测量中估计动脉扩张幅度。在连续测量中，回波跟踪（0.023mm）的相邻心动周期 ML 扩张幅度的标准差显著小于 B 型（0.036mm）或 M 型（0.074mm）方法（p>0.05）。然而，在同一受试者的非连续测量之间，回波跟踪（0.076mm）和 B 型（0.073mm）方法的 ML 扩张幅度标准差相似。傅里叶级数模型和三次平滑样条略微降低了 B 型和 M 型扩张幅度的标准差，但也降低了平均幅度估计值。基于这种方法的相对比较，我们得出结论，尽管回波跟踪在连续测量中提供了高分辨率，但颈动脉扩张非连续测量的重现性，回波跟踪并不比从 30Hz B 型图像获得的更好。
