Добавил:
kiopkiopkiop18@yandex.ru t.me/Prokururor I Вовсе не секретарь, но почту проверяю Опубликованный материал нарушает ваши авторские права? Сообщите нам.
Вуз: Предмет: Файл:
Ординатура / Хирургия / Библиотека им академика М.И. Перельмана / Книга_5772_Библиотеки_им_академика_М_И_Перельмана.pdf
Скачиваний:
0
Добавлен:
30.08.2026
Размер:
58 Мб
Скачать
Physical and Technical Principles of Color Doppler Sonography
The function g(t) is the complex conjugate of g(t). Both func­tions are superimposed and multiplied by each other, and their product is integrated from –to +.
This mathematical folding process yields the value of the autocorrelation at time t. One of the first practical applications of autocorrelation was in electrical engineering to filter out very weak signals from noise and measure them. In ultrasound likewise, autocorrelation can determine the mean frequency shifts of extremely weak blood-flow echoes, whose amplitudes are up to 40 dB fainter than B-mode echoes, against a back­ground of noise. To increase sensitivity at low signal-to-noise ratios, the measurement is repeated several times per scan line. The resulting correlation vectors are added to give the amplitude-weighted temporal mean value of the phase differ­ence, from which the mean velocity can be computed. As a re­sult, the time needed to generate a color image line is several times that needed to produce a gray-scale line. When there is a reasonably high signal-to-noise ratio as in cardiological stud­ies, the number of samplings can be reduced to achieve a higher frame rate. The velocities calculated by autocorrelation are color-encoded and superimposed on the B-mode image (Fig. 1.
13).
1
Other color-encoding techniques are autoregression, the maximum entropy method (MEM), and the time-domain or high-frequency cross-correlation technique mentioned earlier.
Autoregression. Autoregression (AR) techniques are a spin-off of radar technology. Unlike autocorrelation, which analyzes successive pulses at designated time intervals, the time inter­vals are either varied or, in the case of MEM, randomly select-
7
ed
. In practice, the theoretically higher accuracy and resolu­tion of autoregression techniques have been offset by an un­favorable signal-to-noise ratio.
Cross-correlation. The cross-correlation technique compares the high-frequency signal of a pair of echo wavetrains to deter­mine the velocity-related position change of characteristic echoes. An advantage of cross-correlation is its ability to detect even high velocities
8
. The need to have a signal-to-noise ratio higher than 6 dB has limited the routine abdominal use of this method, which can be used only for superficial vessels and in thin patients. Haerten
5
and Liu9have summarized the advan-
tages and disadvantages of the techniques.
Color encoding. It is customary to use a red/blue color encod­ing format in which the color value indicates flow direction and the color brightness indicates flow velocity. There are color encoding maps that are particularly well suited for analyzing slow flows and providing good color fill. In some color-flow maps, green pixels are added to signify variance in flow veloci­ties, indicating turbulence.
12
Fig. 1.13 Longitudinal color duplex scan of the common carotid artery. The white frame outlines the color window in which the mean flow velocity is analyzed for each pixel and superimposed on the B­mode image in color-encoded form. The color scale on the left indi­cates the direction of color encoding (red = flow toward the trans­ducer, blue =flow away from the transducer). Areas devoid of flow are shown in gray scale. The advantage of color encoding is evident when compared with the pureB-mode image just to theleft of the color win­dow: The hypoechoic wall thickening, which is clearly visible in the color duplex image, contrasts poorly with the clear lumen in the B­mode image.
Relevant system settings: B-mode center frequency: 7.2MHz Doppler frequency: 5.14 MHz Pulse repetition frequency: 868 Hz Wall filter: F5 (corresponds to 100 Hz)
Motion detector. In some systems the scan data pass through a motion detector prior to autocorrelation in order to separate the essentially stationary tissue echoes from vascular echoes in the B-mode and color images and to eliminate motion artifacts. This filter can accurately discriminate between color-flow and gray-scale information in the presence of slow flows.
Time averaging (ensemble size). The accuracy of flow-data sampling in color Doppler can be increased by placing a num­ber of transmitted pulses (4–20) at the same site for each color line and performing autocorrelation on all of the sequentially acquired data. Usually the sonographer can set the “time aver­age” to determine the number of samples that are acquired.
Time averaging can increase color sensitivity (improved signal-to-noise ratio for color Doppler), but the higher number of transmitted pulses per color line prolongs the sampling time and reduces the frame rate.
Power Doppler Imaging (or Energy Mode)
Whereas color Doppler involves the direction-dependent de­tection and display of blood flow, power Doppler supplies in­formation on the overall quantity of blood flow. The quantities of the flow components calculated by the correlator are de­tected without regard for direction. They are squared, inte­grated over time, and displayed in a color-encoded format. In­tegrating all the signals significantly improves sensitivity, making it possible to image not just the flow in major vascular trunks but also tissue perfusion (Fig. 1.
Power Doppler is also called transparent energy mode (TEM). This term indicates that power information is added to the B-mode image in such a way that, as in color duplex sono­graphy, it does not mask the B-mode information in areas devoid of flow and leaves that information unchanged. This al­lows for better anatomical orientation.
14).
Fig. 1.14 Comparison of B-mode and power Doppler, illustrated for renal perfusion.
kidney in longitudinal section. There is normal differen-
tiation of the parenchyma and central echo complex.
the branch points of the segmental and interlobar arter­ies could be identified, but only power Doppler can demonstrate parenchymal perfusion. There is no evi­dence of a complete or segmental infarction.
The advantages of the older one-dimensional techniques are their relatively low cost and availability. On the negative side, their use is time-consuming, especially in abdominal studies, and it is not possible to evaluate the perfusion of tissues and tumors.
plex sonography are Doppler techniques, the color information is angle-dependent, much like the information in spectral Doppler analysis. For a constant flow velocity in a vessel, the color shading and intensity can vary depending on the angle between the Doppler beam and the vessel axis. For example,
when a straight, uniformly perfused vascular segment is scanned with a convex transducer array, the varying angles at
which the beams intercept the vessel lead to a color reversal accompanied by a (narrow) black area at the site where the beam is perpendicular to the vessel axis (Fig. 1.
15 a).
Duplex and Color Doppler Sonography
Although the power mode is also a Doppler technique and therefore must be angle-dependent in principle, a signal can still be acquired even when the beam–vessel angle is 90 (Fig. 1.
15 b). This is because the power spectrum has sufficient
bandwidth to detect and integrate all nonzero spectral com­ponents. There are always transducer segments that insonate the pixel volume at an angle other than 90⬚ and can thus con- tribute to the power Doppler image (see Fig. 1. hand, color Doppler information represents only the mean frequency, and so a 90incident beam does not yield a color­flow signal. As a result, power Doppler is largely independent of the beam–vessel angle and is not affected by aliasing, since the direction of the velocity does not enter into the calculation.
viewers, when the spokes in the wheels of a passing stagecoach appear to rotate backward. Pulsed Doppler, with its pulse repe­tition frequency, is analogous to the frames of a movie camera:
With a time delay of T between frames, the viewer cannot tell
19). On the other
Fig. 1.15 Comparison of color duplex sonography and power Dopp­ler: angle dependence.
color Doppler mode shows typical angle dependence of color encod­ing: blood on the left side is flowing away from the transducer and is encoded red, while blood on the right side is flowing toward the trans­ducer and is encoded in the “opposite” color, blue. The blood on the
far right appears light blue because the vessel is scanned at a small Doppler angle relative to the transducer (“pseudojet”). Since the beam-vessel angle is 90at the center of the vessel, no flow is de-
tectable and that area appears black (“pseudo-occlusion”).
13
entire vessel is uniformly filled with color.
Physical and Technical Principles of Color Doppler Sonography
whether the wheel has turned, say, 225forward or –135 backward (Fig. 1.
16). In any sequence of images, the human
brain always interprets the motion as minimal. Thus it will in­terpret rotation as occurring in reverse if the amount of rota­tion between two images is greater than one-half the period of the rotation, i.e., 180. Similarly, when the FFT processor in an ultrasound system calculates the Doppler frequencies from the sampled values of the demodulated Doppler signals at the beat frequency T = 1/PRF, it does so under the assumption of mini­mal frequency values. For example, if the actual frequency is
3
/2 PRF, it will be interpreted and displayed as an aliasing
frequency of
1
/2 PRF.
Nyquist limit. The highest frequency that can be detected without aliasing, called the Nyquist limit, is equal to one-half the PRF in both flow directions in directional Doppler tion 9).
f
ⱕ1/2 PRF (Nyquist limit) (9)
max
1
(equa-
Baseline shift. When aliasing occurs in a Doppler spectrum, positive frequencies above the Nyquist limit are displayed as negative frequencies at the bottom of the spectral trace (Fig. 1.
17). In color Doppler, this “wrap-around” effect appears
as a zone of bright color reversal. This can be corrected in su­perficial vessels by increasing the PRF (Fig. 1.
17, bottom) and in
deeper vessels by shifting the baseline downward or upward (Fig. 1.
18). This can double the range of detectable velocities
both in the Doppler spectrum and in the color-flow image, pro­vided the flow direction is not a concern. The maximum meas­urable velocity can be determined from the considerations on one-dimensional techniques and from the Doppler formula. The Doppler formula yields information on the maximum flow velocity that can be detected without aliasing. This limiting frequency is equal to the pulse repetition frequency PRF. The
14
a
b
1.5
1.0
0.5 0
–0.5 –1.0 –1.5
Fig. 1.16 Principle of aliasing. a PW Doppler and the individual images in color duplex sonography are analogous to the individual frames in a movie camera. If the sam­pling time (T) is too long, the viewer cannot tell whether the wheel has rotated 245forward or 135backward. Since the brain always inter­prets motion in image sequences as being minimal, more than 180of rotation (between two frames) will appear as reverse motion on the film. b Aliasing also occurs in the recording of Doppler spectra. When the frequency to be measured (red curve) is sampled (arrows) at a rate of less than one-half the pulse repetition frequency, the underlying frequency cannot be unambiguously identified. The sampled points trace out a curve of lower frequency (blue curve). This causes a “wrap­ping around” of the spectral trace in the Doppler spectrum, as high positive frequencies that cannot be displayed above the baseline are truncated and appear as negative frequencies at the bottom of the spectrum. Aliasing in colorDoppler causes the flow velocitiesto be dis­played in the opposite color. When the sampling rate is equal to at least one-half the pulse repetition frequency (the Nyquist criterion), the curve can be plotted at the correct frequencies.
T
225°
–135°
Fig. 1.17 Aliasing in color and spectral Doppler, before and after ad­justment of the pulse repetition frequency (PRF). a With a PRF of 2256 Hz, aliasing appears as a color reversal in the color-flow image andas a wrap-around of the truncated spectralpeaks in the Doppler spectrum. b When the PRF is increased to more than 5000 Hz for color Doppler or 4000 Hz for spectral Doppler, a normal, artifact-free color and spec­tral display of flow information is obtained.
Doppler parameters
Duplex and Color Doppler Sonography
f Mean Doppler frequency σ2 Variance
P Power A Doppler signal amplitude
a
Noise
–PRF/2
Angle dependence
Θ=90°
P=P1+P2>0
f =0
2
A
0
2
A
f v· cos Θ
f v· cos Θ
b
Θ≠90°
f PRF/2
Wall filter
2
σ
P
Physical and Technical Principles
c
Fig. 1.18 Aliasing due to incorrect placement of the spectral base­line. a Severe aliasing caused by setting the baseline too high. Almost half of the spectrum has been truncated and displayed below the baseline. b After an initial baseline shift, most of the spectrum is already dis­played normally. c When the baseline is lowered further, almost theentire PRF is availa­ble for sampling, and the Doppler spectrum is free of artifacts. Reverse
flow components cannot be identified, however.
maximum detectable velocity, then, decreases with increasing depth, and it increases as the transmission frequency is re­duced.
Because all of the flow components are integrated over a time interval, power Doppler is very sensitive in detecting per­fusion but is also much more susceptible to relative move­ments of the organs or transducer and therefore requires con­siderable user skill.
Comparison of spectral Doppler, color Doppler, and power Doppler. Therelationship among spectral Doppler, color Dopp-
ler, and power Doppler is reviewed in Fig. 1.
19. The figure
–PRF/2 0 f PRF/2
Fig. 1.19 Doppler parameters in the amplitude–frequency curve. In both diagrams, the squares of the amplitudes of the flow components are plotted over the frequency axis. a Doppler parameters in context. Color Doppler presents the mean Doppler frequency (
f). The variance (
2
σ
) is a measure of turbulence, corresponding to the bandwidth of the spectrum. Power Doppler is based on the area under the curve. Spectral Doppler displays all
frequency components over the time axis.
b Special case of a 90insonation angle. In color Doppler, a mean flow
velocity of zero is calculated in both directions from the point where the beam is perpendicular. In power Doppler, on the other hand, the various directional components are added together, still allowing flow
detection to occur.
shows the amplitudes (squared) of all the Doppler frequencies that can be measured for each sample volume and time inter-
val:
Spectral Doppler displays all of the amplitudes in the range gate in a time-varying, brightness-modulated spectral trace.
In color Doppler, only the mean frequency shift f = kx cos Θ at each point in the image is color-encoded. The variance is a measure of the bandwidth of the Doppler spectrum, i.e., how diverse the flow velocities are at the sampling site.
Power Doppler encodes each pixel with the sum of all ampli­tudes, i.e., the area under the curve (amplitudes squared) averaged over time.
15
Physical and Technical Principles of Color Doppler Sonography
Technical Aspects of the Examination and
Equipment Settings
Numerous technical aspects of the ultrasound examination and system settings must be considered in the one- and two­dimensional Doppler techniques in order to optimize the ex­amination and avoid diagnostic errors.
Insonation angle. The most important scan parameter is the angle between the Doppler beam and the vessel axis. Because the cosine of this angle is used in the Doppler formula, the in­sonation angle should be kept well below 90⬚ during the ex- amination. When the angle is perpendicular, a Doppler signal is not recorded. To determine the velocity v from the Doppler frequency, it is necessary to measure the angle in the B-mode image and perform an “angle correction.” The accuracy of the angle correction depends greatly on the size of the beam – ves­sel angle. Table 1. velocity is minor at small angles but becomes substantial at large angles. An accurate calculation of flow velocity is possible only when the beam – vessel angle is small (see Analysis of Doppler Information and Artifacts, p. 19).
1
Transmission frequency. Selection of the transmission frequency for flow detection depends on the sampling depth and the anticipated velocity. As described under Penetration Depth (p.5), high ultrasound frequencies are strongly at­tenuated with depth. This is particularly true of the Doppler signal, which is considerably weaker than the intensity of the echoes reflected from vessel walls and organs. At the same time, there are conditional upper and lower limits to the de­tectability of Doppler frequencies for technical reasons. Be­cause the Doppler shift frequency is directly proportional to the velocity to be measured, it is advantageous to use a low transmission frequency for the measurement of high veloci­ties. By the same token, a high transmission frequency should be used for low velocities.
Power output and receiver gain (Doppler or color gain). These settings should be adjusted so that color completely fills the vessel lumen with no extraluminal “blooming” and the dis­played Doppler spectrum is free of noise and superimposed signals.
2 shows that the effect on the angle-corrected
Velocity scale. In both spectral and color Doppler imaging, the scan frequency or velocity scale should be set for the antici­pated range of velocities. Setting the scale too high makes it more difficult to detect and analyze the signals; venous signals in particular may go undetected. If the flow velocity in the ves­sel exceeds the maximum value set on the velocity scale (PRF) in one flow direction, this higher velocity will be misinter­preted as a high velocity in the opposite direction. One possible remedy is to increase the PRF. If this is no longer possible, the beam–vessel angle can be deliberately increased, or the scan­ning frequency may be decreased as described above. When aliasing occurs in the color-flow image, the color changes are alwaysin the range of higher velocities (e.g., white or yellow).If the flow direction is actually changed owing to turbulence or flow reversal, the color changes run through black (Fig. 1. Power Doppler can mask these phenomena, which are often helpful diagnostically. On the other hand, the angiogram-like appearance of the vessels in power mode can aid in the detec­tion of vascular lesions (Fig. 1.
Wall filter and motion-artifact filter. These filters suppress troublesome pulsations and flash artifacts, but they can also filter out diagnostically relevant low frequencies causing slow flows to be missed (Fig. 1. priority (color balance). If the priority is set too high in the B­mode image, flow may not be detected (see Artifacts and Pitfalls, p. 20).
Time averaging, color box, line density. Parameters such as time averaging, which increases sensitivity, and the line den­sity reduce the frame rate. Tokeep the color frame rate in an ac­ceptable range, it is good practice to limit the color-flow infor­mation to a color box, which should be of minimal size in its lateral extent and depth. In some examinations, such as fetal heart imaging, it is also helpful to reduce the color line density (at the cost of decreased color spatial resolution) in orderto en­sure adequate temporal resolution for cardiac imaging.
Another way to increase the color frame rate is to use a sys­tem that offers parallel processing of the echo signals from a transmitted pulse (see New Technical Processes and Ap­proaches, p. 24). With this type of system, the information from multiple adjacent scan lines can be processed simul­taneously.
Recommendations on key equipment settings for CW, PW, and color Doppler examinations are summarized in Table 1.
20).
21). The same applies to color-write
20).
3.
16
Table 1.2 Effect of beam-vessel angle on angle correction errors
Beam-vessel angle Correction factor
Θ)
(1/cos
30 1.15 3% 45 1.41 6% 60 2.00 9% 72 3.24 15 % 75 3.86 21 % 80 5.76 30%
Angle correction errors are calculated by assumption of a detection accuracy of 3%
Correction error
Duplex and Color Doppler Sonography
a
Fig. 1.20 Comparison of turbulence and aliasing in color duplex and power Doppler imaging. a Color duplex image of a vascular bifurcation shows local wall thick­ening and zones of turbulence. Because the lower vessel branches off at an angle from the main vessel, the change of angle between the color box and vessel axis causes apparent flow acceleration in the lower vessel. Color aliasing is recognized by noting that the color
a
b
change is in the range of the brighter colors on the velocity scale (e.g.,
white or yellow). With a true change in flow direction due to tur­bulence or flow reversal, the color would by running through black. b Power Doppler shows a homogeneous vessel lumen in which local
wall thickening at the origin of the upper vessel appears dark. Tur­bulence and aliasing are not visualized, but the local wall lesion is clearly appreciated in the angiogram-like view.
b
Fig. 1.21 Effect of wall filter setting on frequency detection in spec-
tral Doppler. a Normal spectrum of an artery with high residual diastolic flow and low resistance in the distal vascular bed. Wall filter is set at 25 Hz. b When the wall filter is increased to 225 Hz, velocities in the range
from 0 to10 cm/s arenot displayed (angle correction from the B-mode image is constant for all settings in a c). c When the wall filter is further increased to 400 Hz, velocities in the range from 0 to 20 cm/s are lost, creating an apparent spectral pattern of absent end-diastolic flow. Tumor and tissue perfusion and slow
venous flows may go undetected.
Physical and Technical Principles
c
17
Physical and Technical Principles of Color Doppler Sonography
Table 1.3 Recommended system settings for CW, PW and color Doppler examinations
Settings CW PW Color
Doppler
Transmit power, transmit intensity + + + The necessary transmit power depends on the examination conditions
Receiver gain + + + Set the gain to produce a spectrum with minimal noise and no superim-
Pulse repetition frequency (PRF), velocity scale
Baseline setting + + + Adjust so that the range of measurable frequencies is optimally utilized.
Wall filter and flash-artifact filter + + + Troublesome wall pulsations and motion artifacts are reduced, but low
+ + + Set the PRF to fully utilize the range of measurable frequencies. A PRF of
1
Color priority + Postprocessing parameter that controls the balance between gray-scale
Settings, problems, and optimization
(sound absorption) and on depth. The power setting for obstetric exami­nations should be as low as possible.
posed signals or a color-flow image in which a normal vascular segment is completely filled with color and there is no color blooming or superim­posed color noise.
1000Hz is recommended for veins, 1500–3000Hz for arteries. May have
to increase the PRF if aliasing occurs, or may have to selectively increase the Doppler angle for deep abdominal scanning. In some cases it may
also be necessary to reduce the scanning frequency (see equation 6).
Shifting the baseline is helpful if aliasing occurs.
velocities may not be detected. Filter setting of 50–200 Hz is recom-
mended for arteries, ⬍100Hz for veins. Use of other filters such as mo-
tion and flash-artifact filters can reduce sensitivity, especially for the de­tection of slow flows.
echoes and color pixels. If the color priority is set too low, it can suppress
flow signals in the B-mode image. To search for thrombosis or reduce
noise, it should be set for slight gray-scale priority.
18
Line density for B-mode/color flow (synonym: spatial resolution)
Temporal resolution + Sets the number of transmitted pulses used to compute an ultrasound
Correlation (persistence) + Affects the composition of color information by the weighted addition of
Spatial averaging + One- or two-dimensional filtering of the color image based on the sur-
+ Improves the lateral resolution of the B-mode or color image at the cost
of a reduced frame rate. If fast motion (e.g., fetal heart) or high velocities are to be detected, the line density should be reduced.
color line or the number of acoustic lines used in autocorrelation. In­creasing temporal resolution improves the signal-to-noise ratio and sensi-
tivity. The disadvantage is a reduced frame rate.
old and new information. The higher the number, the more old informa-
tion is retained. Time averaging the color pixel values helps to image all flow in the vessel lumen but slows the color presentation. The disadvan­tage of a high setting is less appreciation of temporal dynamics.
rounding color pixels or color information. Reduces color noise without affecting the frame rate, but can lead to nonvisualization of very small
vessels.
tain a clear spectral window below the trace. For two-dimen-
Analysis of Doppler Information and Artifacts
sional color flow imaging, the PRF, gain, and any filter settings should first be adjusted in a normal vascular segment so that
Optimizing the Examination and Analyzing the Doppler Information
Settings before the start of the examination. After selecting
the proper transducer for the desired information and the
color uniformly fills the vessel lumen. We recommend the fol­lowing technique. After identifying the vessel and angling the color box in proper relation to the vessel axis, adjust the power setting, PRF, and gain until a noisy color image is obtained.
Then lower the gain setting until the color noise is gone. proper frequency for Doppler interrogation, the examiner should adjust and optimize the settings listed in Table 1. obtain a well-defined spectrum that is free of noise and alias­ing (Fig. 1.
22). For examinations of the fetal aorta and maternal
arteries, the power and gain settings should be reduced to ob-
3 to
Information. Information on flow obstructions at the sampling
site itself and on conditions in the proximal and distal vascular
beds can be derived from the signal frequencies, which are a
measure of flow velocities; from the amplitudes, which are
Duplex and Color Doppler Sonography
a
Fig. 1.22 Dependence of the Doppler spectrum on resistance in the distal vascular bed. a Spectrum of an artery with high resistance in the distal vascular bed.
This resistance leads to a rapid decline of flow after systole, with a reverse flow component appearing in diastole. All the red blood cells move at approximately the same velocity during systole, creating a narrow frequency band with a clear “sonic window.“ b Spectrum of an artery with low resistance in the distal vascular bed,
with persistent residual diastolic flow. c Venous signal from the inferior vena cava. Unlike the bandlike spec-
trum of peripheral veins or the portal vein, the superior and inferior
vena cava show cardiac modulation of their flow, which can easily be confused with an arterial flow pattern.
proportional to the number of moving blood cells; and from the time changes over each cardiac cycle. Color Doppler imag­ing and spectral analysis are considered to be complementary in the evaluation and interpretation of Doppler information.
When applied together, both color and spectral Doppler can be a valuable source of qualitative, semiquantitative, and quanti­tative information.
b
Physical and Technical Principles
c
Analysis of the Doppler Spectrum
Waveform. The Doppler spectrum can be described in terms of the spectral bandwidth and the shape of the waveform (en-
velope curve), which is basically characterized by its steepness, peak systolic velocity,and end-diastolic velocity. Moving left to right, the trace begins its upstroke at the start of systole and concludes at end diastole. One or two cardiac cycles are always evaluated for spectral analysis. The shape of the spectrum de­pends on local findings and on the organ that is supplied by the sampled vessel (Figs. 1.
Spectral bandwidth. If all of the blood cells are moving at the same velocity, the spectrum displays a narrow frequency band
with a typical clear “sonic window” below the trace (Fig. 1.
This is a normal finding in early systole for all larger vessels.
Vessel wall irregularities produce turbulence, causing initial spectral widening or f illing-in of the spectral window. Starting at about 50% stenosis, a jet develops owing to local flow accel­eration. With proper angle correction, this acceleration can be identified in the Doppler spectrum.
Pulsatility index (PI). Several techniques have been devised for analyzing the spectral Doppler waveform. Gosling first to describe an index for analyzing pulsatility. This is a
22,1.23).
4
was the
22).
Fig. 1.23 Doppler spectrum with parameters for spectral analysis (S/D, RI, PI). This spectrum is from a vessel with a low-resistance distal bed. After the operator manually marks the starting and end points of
two cardiac cycles, most scanners will automatically calculate the key flow parameters and Doppler indices. While the velocity measure-
ments are angle-dependent and require corresponding angle correc-
tion in the B-mode image, all of the indices listed below are indepen­dent of the beam anglebecause in each case the correction factor can­cels out of the equation.
S: peak systolic velocity (here: 55.3 cm/s) D: end-diastolic velocity (here: 32.3 cm/s)
Tav: time-average peak velocity (here: 39.4 cm/s) Pulsatility index PI = (S–D)/TAV (here: 0.59) Resistance index RI = (S–D)/S (here: 0.42) Stuart index S/D (here: 1.72) Maulik D/TAV
19
Physical and Technical Principles of Color Doppler Sonography
measure of the difference in flow between systole and dias­tole during the cycle; it was first determined by calculation from the Fourier analysis. A simplified version is the Pulsatil­ity index (PI). This is calculated by dividing the difference be­tween the peak systolic and end-diastolic frequencies (D) by the time average of the maximum frequency shift (TAV) (Fig. 1.
23).
Resistance index (RI). Pourcelot introduced the resistance index, an angle-independent measure of pulsatility in which S describes the peak systolic frequency and D the end-diastolic frequency. Many sonographers use simplified indices like that of Stuart and Drumm frequency S and end-diastolic frequency D. Maulik proposed the ratio D/TAV, in which the maximum end-diastolic frequency is normalized to the mean Doppler shift frequency of the outer envelope curve.
S/D ratio, A/B ratio. Other parameters for waveform analysis were introduced in 1983 by Campbell et al.
17
al.
. These were based on curve fitting and on descriptions of steepness, maximum frequency, and other defined slopes. But from the many indices that have been devised for semiquanti-
1
tative analysis,only the S/D ratio (also called also the A/B ratio), the resistance index (RI), and the pulsatility index (PI) have be­come widely used. The formula RI = 1 – 1/(S/D) is used to con­vert between the RI and S/D ratio.
15
, which is the ratio of the peak systolic
2
and Thompson et
and can be confirmed and described semiquantitatively by an­alyzing the Doppler spectrum. Apparent changes are caused by changes of insonation angles within a vessel (Figs. 1. and by the occurrence of aliasing when the PRF is set too low. Other pathological changes are seen in association with mal­formations, stenoses, and occlusions.
Stenoses. The oldest indication for Doppler scanning was the detection of stenoses. These sites are recognized in color du­plex sonography by typical local flow acceleration, which is usually manifested by a bright color jet or aliasing (Fig. 1. Measurements of maximum flow velocity help to confirm and quantitate the stenosis. Velocities of 140–180 cm/s are suspi­cious, while higher values confirm that a stenosis is present. Additional criteria include eddy flow located within and just past the stenosis and poststenotic dilatation, which is often visible in the B-mode image. Other, indirect criteria are the presence of distal poststenotic damping and increased pre­stenotic resistance, which causes a “flickering” of the color pixels. Poststenotic damping causes a loss of pulsatility in the color image and decreased brightness modulation in the spec­trum.
Whereas the diagnosis of an intra-abdominal stenosis is generally based purely on hemodynamic criteria in the color­flow image, the color duplex scanning of superficial vessels with a high-resolution transducer can additionally define the morphology of the stenotic lesions (Fig. 1.
25).
15,1.24)
25).
20
Quantitative information. Measurements of the Doppler shift frequencies with angle correction supply quantitative infor­mation on flow velocity and also on volume flow when the cross-sectional area is considered. Owing to errors in angle de­termination, the effect of filters, inaccuracies in estimating the true center frequency, and problems in determining the exact vascular cross section used in calculations, the importance of volume measurements has dwindled in favor of angle-inde­pendent semiquantitative parameters (PI, RI, S/D).
Analysis of Color Information
While the Doppler spectrum portrays the time course of flow at a selected sampling site, the color information supplied by two-dimensional Doppler techniques describes the flow con­ditions that prevail within a volume element at a given point in time. Color changes, like spectral shape, are influenced by physiological and pathophysiological states in the sampled vessel and in the proximal and distal vascular beds. Besides qualitative information (flow: yes/no), color image analysis fo­cuses on the time course of color saturation and the color dis­tribution:
Reverse flow components. Physiological changes in flow and flow direction occur at vessel branch points such as the carotid bifurcation, consisting of flow reversal or a reverse flow com­ponent in high-resistance vessels such as the aorta and the iliac arteries (Fig. 1. bral artery and umbilical artery (in the second and third trimesters) show unidirectional flow that persists even in dias­tole, appearing as slightly darker residual flow. The presence of a reverse flow component in these vessels is a pathological sign
24). Low-resistance vessels like the middle cere-
Thromboses and arteriovenous (AV) fistulae. Color Doppler may also demonstrate voids in the color-filled lumen signify­ing thrombosis or a coarse, mosaic-like vibration artifact (“confetti sign”). This may be an accompanying feature of a high-grade stenosis, or it may signify an AV fistula (Fig. 1. The lesions are differentiated by performing a spectral analysis and checking for secondary fistula signs such as increased flow in the feeding artery and/or an “arterialized” spectrum of the draining vein.
Tumors. Thedetection of blood flow aids in identifying tumors, and the specific blood flow pattern can advance the differential diagnosis. Newer approaches, such as calculating the inflow rate of ultrasound contrast medium in tumors or in the kidney, can furnish quantitative information on tissue perfusion and renal function.
26).
Artifacts and Pitfalls
Artifacts and possible diagnostic errors can result from im­proper settings or from unexpected physical or equipment limitations. Some artifacts cause loss of sensitivity, while others give a false-positive indication of flow.
Absent (False-Negative) Flow Detection
Besides the power and gain settings, the detection of flow de­pends on the wall-filter and motion artifact-filter settings (Fig. 1.
21) and on the selected PRF. Other system-specific pa-
rameters are often underestimated in their significance. These include the color priority, which controls the representation of color in the B-mode image. Setting the color priority too low
Duplex and Color Doppler Sonography
ab
Fig. 1.24 Difference between physiological flow reversal, reverse
flow component, and aliasing.
a Physiological flow reversal in the carotid sinus. Zones of eddy flowat
the carotid bifurcation are encoded in blue.
b Reverse flow component in a vessel with high diastolic resistance.
The cranially directed flow is encoded in red. The reverse flow com­ponent, encoded in blue, is bordered by a dark zone representing the zero baseline (compare with spectrum in Fig. 1.22 a). c Aliasing due to apparent flow acceleration (“pseudojet”). The cause is the smaller Doppler angle that results from the curvature of the ves­sel. When angle correction was applied, a normal flow velocity was de-
tected in all segments.
Physical and Technical Principles
a
Fig. 1.25 Color duplex images of stenoses in the neck and in abdomi­nal arteries. a Longitudinal scan through the liver and aorta shows a high-grade stenosis of the celiac trunkwith aliasing and moderate poststenotic di­latation. The aorta and the origin of the superior mesenteric artery ap­pear normal. b Transverse scan through the vena cava (encoded blue), aorta, bra­chiocephalic trunk, splenic artery, and occluded common hepatic
b
c
c
artery. The insonation angle is large, but scans at all angles confirmed
the absence of flow signals.
c Seventy-percent stenosis at the origin of the internal carotid artery,
with a short jet phenomenon. Angle-corrected Doppler measure­ments showed arise in the peak frequency tomore than 400 cm/s. The cause is a tubular wall thickening 1.5 cm long with a proximal filiform constriction.
21