Page 80 - Libro vascular I

Chap-06.qxd 29~8~04 14:41 Page 71
FACTORS THAT INFLUENCE THE DOPPLER SPECTRUM
71
of error in vessel diameter measurement are dis- cussed below.
Image resolution
The ability to image an object is dependent on the resolution of the scanner, as described in Chapter 2. The resolution along the axis of the beam is better than that across the image (i.e., the lateral resolu- tion). The axial resolution is of the order of the wavelength of the ultrasound. For example, the wavelength of a 3 MHz transducer is 0.5 mm, whereas the wavelength of a 10 MHz transducer is 0.15 mm, the latter therefore providing more accu- rate distance measurements. Lateral measurements are much less accurate, as a result of the poorer image resolution and reduced image quality due to the beam being parallel to the vessel wall. The ves- sel diameter is especially difficult to measure in the presence of disease.
Calliper velocity calibration
Accurate diameter measurements rely on correct calliper velocity calibration. Most scanners assume the mean velocity in tissue to be 1540 m/s; how- ever, the velocity of sound in blood is actually 1580m/s. This results in a systematic underesti- mate of the order of 2.6% in diameter measurement, leading to a 5% error in cross-sectional area.
Variable vessel diameter
The arterial diameter is not, in fact, constant but varies during the cardiac cycle due to the changing pressure within the vessel. This means that a single measurement of the diameter may not be repre- sentative of the mean diameter. It has been shown that vessel wall pulsatility may result in up to a 10% change in vessel diameter between systole and dias- tole. This cyclical variation in diameter will lead to errors in volume flow estimation, but it may be reduced by taking several diameter measurements and finding a mean value. Ideally, an instantaneous diameter measurement should be multiplied by the instantaneous mean velocity to obtain a more accu- rate volume flow measurement, but this technique is not currently available on commercial ultrasound scanners.
Noncircularity of the vessel lumen
The calculation of cross-sectional area from the diameter measurement assumes that the vessel lumen is circular, which may not be the case, espe- cially in the presence of disease.
Errors in measuring TAV
Incomplete insonation of the vessel will lead to an underestimation of the proportion of slower moving blood at the vessel wall, which in turn will lead to errors in the mean velocity measurements. For example, if a Doppler recording is obtained from a vessel with parabolic flow using a narrow beam (as shown in Fig. 6.2A and B), the high- velocity flow in the center of the vessel will be ade- quately sampled, but a large proportion of the slower moving blood at the vessel wall will not be detected. When the mean velocity is calculated from the spectra, this will be an overestimate of the true mean velocity due to the undersampling of the flow at the lateral edges of the vessel. This is true even if the sample volume is set to cover the near and far walls of the vessel as the out-of-imaging plane flow will not be sampled. Incomplete inso- nation of the vessel can lead to errors of up to 30% in the TAV (Evans & McDicken 2000).
Alternatively, the mean TAV can be estimated from the maximum TAV if the flow is measured at an adequate distance from geometric changes (e.g., bifurcations or stenosis) and the shape of the flow profile in the vessel is known. If there is a blunt flow, the maximum velocity will be equal to the mean velocity across the vessel. However, if the velocity is parabolic then the maximum veloc- ity will be twice the value of the mean. One advan- tage of the maximum velocity measurement is that it is not affected by the width of the beam, provided the beam passes through the center of the vessel.
If the wall thump filter is set too high, the low- frequency signals from the slower moving flow will be removed, and this would lead to an overesti- mate in the mean velocity. Aliasing would lead to underestimation of the mean velocity due to the incorrect estimation of the high frequencies present within the signal. The presence of high-amplitude noise will bias the estimate of the mean velocity,