echoes of the original sound beam initially created by placing the ultrasound transducer in contact with

the tissue. The B-mode component of the duplex machine analyzes the strength (intensity) and origin of

the reflected echo. The Doppler component analyzes shifts in frequency of the original sound wave

produced by the transducer that result when the ultrasonic wave encounters moving reflectors.

B-Mode Ultrasound

As a sound wave passes through tissue, its strength moving away from the transducer depends upon the

degree to which the beam is attenuated, scattered, and reflected. The strength of the reflected echo

depends in part, upon relative differences in acoustic impedance between two different media. Major

differences in acoustic impedance result in reflection of a large proportion of the sound beam back to

the transducer, whereas small differences result in comparatively little reflection and continued

propagation of the beam through the tissue.

In B-mode ultrasound, the strength of the returning echo is reflected in the brightness of the

individual pixels comprising the ultrasound image. This is the ultrasound grayscale and the resulting

image is termed a grayscale image. (As we will see below, the primary purpose of the grayscale image in

duplex ultrasonography is to aid in properly positioning the Doppler sample volume.)

Very bright pixels represent sites of large differences in acoustic impedance between media, whereas

less dramatic differences are represented by proportionally less bright pixels on the B-mode image. Thus

gallstones, which differ dramatically in acoustic properties from soft tissue, result in strong echoes and

proportionally brighter pixels on an ultrasound image, whereas blood, which differs little from soft

tissue in acoustic characteristics, frequently cannot be distinguished from soft tissue in a grayscale

image.

The strength of the reflected echo will also be dependent upon the strength of the sound beam at the

point of its reflection. Grayscale images do not show the percentage of the beam reflected, but rather

represent the absolute strength of the reflected echo arriving back at the transducer. Thus if the sound

beam is very weak at the site of its reflection, even areas of significant differences in acoustic

impedance will not result in a bright pixel on the ultrasound image.

The strength of the ultrasound beam at any point also depends on the degree to which the beam has

been attenuated as it passes through the tissue. Attenuation of a sound wave as it traverses tissue

weakens the wave and depends upon the tissue traversed and the frequency of the wave. The frequency

of the wave depends upon the frequency of the transducer used to generate the wave (see discussion

above and Equation 87-1). Sound waves resulting from higher-frequency transducers are attenuated

more rapidly than those produced with lower-frequency transducers. Higher-frequency transducers

therefore provide relatively weak echoes to be reflected from a deep structure and thus a comparatively

poor B-mode image compared to a lower frequency transducer.

The linear resolution of an ultrasound image depends upon the ability to focus the beam. Sound

beams emanating from high-frequency transducers can be more precisely focused than those from lowfrequency transducers and thus provide clearer B-mode images of more superficial structures. Because

the carotid artery is superficial, higher-frequency transducers can be used to provide much clearer Bmode images than is possible with deeper vessels such as the aorta, renal, or iliac arteries.

Doppler Ultrasound

In a continuous wave Doppler, a transducer continually emits vibrations into the tissue. Echoes are

therefore continually reflected back to the transducer. A transducer cannot, however, simultaneously

generate and receive an echo. Continuous wave Doppler therefore must have separate transmitters and

receivers to both generate and receive echoes.

1 Duplex devices utilize pulse Doppler. A pulse Doppler uses the same transducer to generate and

receive echoes. Because the speed of sound is relatively constant in tissue, with a pulse Doppler it can

be determined when an echo is generated and when it is received and therefore it is possible to

calculate the depth in the tissue where a reflected echo originated. The sample volume of the duplex

device is thus determined by specifying from what depth one wishes the pulse Doppler to receive

reflected echoes.

The transducer is gated, based on the total time from original sound-wave generation to arrival of the

reflected echo back at the transducer, to only receive echoes from the specified depth of the sample

volume. The desired “position” of the sample volume is determined by the operator of the duplex device

from the B-mode image, and in duplex examination of arterial and venous flow corresponds to

particular points within the lumen of the vessel examined. Because B-mode images and Doppler

2485

waveforms cannot be generated simultaneously, the technologist must continually update the B-mode

image during the course of the examination to insure proper placement of the sample volume.

2 When a sound wave encounters moving reflectors, such as red blood cells, so-called Rayleigh

scatters, within the lumen of an artery or vein, the frequency of the reflected wave changes from that of

the original wave generated by the transducer. Velocity is a vector with both speed and direction

therefore, the magnitude of the frequency shift depends upon the velocity of the moving reflectors and

their angle with the incident sound beam. Because the frequency of the original sound wave is known,

and the frequency of the received echo can be determined by the software of the duplex machine, the

velocity of the moving reflectors (red blood cells) can be calculated provided the angle of the incident

sound beam with the moving reflectors is also known. This relationship is reflected in the Doppler

equation:

Where fr is the received frequency, fo

is the originating frequency, v is the velocity of the reflector, c

is the speed of sound in tissue, and θ is the angle of the incident sound beam with the moving reflectors,

the so-called Doppler angle.

A Doppler angle of 0 degrees is ideal for velocity calculations as the cosign of 0 is 1 and measured

reflected velocity changes are therefore 100% of what actually occurred. However 0-degree Doppler

angles are difficult to obtain as it is not generally possible to look directly down the lumen of a vessel

from a percutaneous approach. Therefore to solve the Doppler equation for the velocity of the moving

reflectors, the Doppler angle must be known. To standardize the results of duplex scanning, it has been

traditionally recommended that examinations be conducted as close as possible to a Doppler angle of 60

degrees. The cosign of 60 is 0.5, therefore at this angle the measured reflected velocity is half the actual

reflected velocity. Errors in velocity calculations secondary to misreading of the Doppler angle are small

as a percentage of the true velocity at ≤60 degrees as cosigns vary comparatively little with

incremental changes in angle below 60 degrees. When the Doppler angle approaches 70 degrees,

changes in the cosign of the angle become more pronounced with incremental increases in angle, and

errors in velocity calculations secondary to improper angle measurement become much more

significant. Small errors in determining the Doppler angle when the angle of insonation is <60 degrees

have little overall impact on the calculation of the velocity of the moving red cells. Indeed, Doppler

angles from 45 to 60 degrees are acceptable for most clinical studies. Most often an angle of 60 degrees

can be easily obtained in carotid artery and extremity artery duplex examinations and, in line with

traditional teaching, 60 degrees is chosen whenever possible as the Doppler angle in such studies. In

examination of intra-abdominal vessels it may be much more difficult to adhere to a 60-degree angle.

As noted above, pulse Doppler both transmits and receives echoes with the same transducer. This dual

function places limits on the frequencies that can be displayed in the spectral waveform. The maximum

frequency that can be displayed is half the pulse-transmitting frequency or pulse-repetition frequency

(PRF), and is known as the Nyquist limit. If frequencies are encountered that exceed the Nyquist limit,

they are flipped and appear on the reverse-flow side of the spectrum. This is termed aliasing. If aliasing

is encountered, the PRF should be increased to increase the Nyquist limit. The PRF can be adjusted by

changing the scale displayed on the right side of the spectral waveform or by adjusting the baseline. If

aliasing still occurs, the operator should consider switching to a lower frequency transducer or to the

use of a continuous-wave Doppler. A continuous-wave Doppler is not affected by aliasing.

Color Flow

A color flow image is produced by assigning color to Doppler shifts. Returning echoes that are not

Doppler shifted are shown in grayscale. The net result is what appears to be color pasted or

superimposed on a grayscale image. This pasting process can in effect obscure the edges of the grayscale

image resulting in obfuscation of the true lumen of the grayscale image limiting somewhat the accuracy

of color images to define stenosis. The hue and intensity of the color are determined by the direction

and magnitude of the Doppler shifts. The technologist can adjust the color settings as desired. Varying

shades of red and blue are generally used to distinguish flow toward or away from the transducer. By

convention, red is most often assigned to arterial flow and blue to venous flow.

When the Doppler angle is 90 degrees, there is no Doppler shift. When this occurs, the assigned color

is black and corresponds to the horizontal black bar serving as the baseline on the color scale that

appears on the right side of a color flow image (Fig. 87-1). Aliasing can be readily recognized on a color

flow image as a mosaic pattern or as a transition from red-to-blue or blue-to-red without an intervening

2486

black line. In areas of actual reverse flow, the blue and red colors are separated by a black border. Just

as with pulsed Doppler, color aliasing can be reduced by increasing the PRF. This is often quite

important, as an extensive mosaic pattern can obscure clear visualization of a stenotic site.

Power Doppler is a variant of color Doppler. Power Doppler, however, assigns no direction of flow to

the color image and velocity information is not calculated. Any detected Doppler shift is colorized

without consideration of the direction of the Doppler shift or the magnitude of the Doppler shift. It is

therefore not subject to aliasing or affected by Doppler angle. Its principle utility lies in the ability to

detect blood moving at very low velocity and in outlining the course of tortuous vessels. It can be quite

useful in detecting flow in small distal veins and flow in arteries distal to a very high-grade stenosis.

It is tempting to try a direct estimate of the severity of a stenosis from the color flow image.

However as discussed above such estimates are probably less accurate than measurements of stenosis

derived from spectral analysis. Color serves primarily as a guide in locating the vessels and selecting

specific sites for examination with the pulse Doppler. The absence of color can indicate extensive mural

calcification and difficulty in obtaining a pulse Doppler waveform.

Figure 87-1. Flow separation with reverse flow along the wall of the carotid bulb as indicated by direct transition from blue to red

with an intervening black line is a normal finding and indicates an artery without significant plaque.

Carotid and Vertebral Arteries

The vascular laboratory began with assessing internal carotid artery (ICA) stenosis. Doppler diagnosis of

ICA stenosis, and arterial stenosis in general is based on duplex ultrasound and focuses on three areas,

the prestenotic region, the stenosis itself and the poststenotic region. Detecting carotid stenosis involves

a combination of spectral analysis, color and grayscale imaging and, in selected cases, power Doppler.

The primary interest is in the detection of increased flow velocities within the area of suspected stenosis

and therefore in most cases spectral analysis dominates in the detection of carotid stenosis. The

exception is in distinguishing ICA occlusion from a very high-grade stenosis where color flow or power

Doppler flow can identify very low flow in the area of the stenosis or distal to the stenosis that may not

be detected by pulse Doppler.

In all cases, the pulsed Doppler and color flow findings should be crosschecked for concordance. If

there is disagreement between the impression obtained with color and pulsed Doppler examinations

(i.e., color Doppler suggests high-grade stenosis but velocities are only moderately elevated), the

findings of both should be reviewed to resolve the discrepancy.

Common Carotid Artery (CCA) Waveforms

3 In the majority of cases, carotid stenosis or occlusion occurs in the proximal ICA at or just beyond the

carotid bifurcation. A normal ICA waveform reflects the low resistance of the cerebral circulation with

high flow at the end of the diastolic component of the waveform. In the presence of a distal ICA

occlusion resistance to flow increases and there is decreased flow in diastole (Fig. 87-2A,B). The

external carotid artery (ECA) supplies the relatively higher-resistance circulation of the scalp and face

and therefore has a low-end diastolic flow component (Fig. 87-2C). The common carotid artery

waveform reflects the fact that normally 80% of CCA flow is directed to the ICA. CCA end-diastolic flow

is therefore generally well above baseline and exceeds ECA diastolic flow (Fig. 87-2D). In the presence

of a very high-grade ICA stenosis or ICA occlusion, outflow is primarily through the higher-resistance

2487

external carotid circulation. The CCA waveform then takes on the high-flow resistance characteristics of

an ECA, with flow to zero, or nearly zero, in end diastole (Fig. 87-3).1 In addition, the PSV and the

overall flow velocity may be substantially lower than normal because of reduced carotid artery flow. By

observing these changes in the CCA, one can reliably predict the presence of high-grade stenosis or

occlusion of the ICA.

Figure 87-2. A: A normal internal carotid artery (ICA) waveform with high diastolic flow (white arrow) reflecting the low

resistance of the cerebral circulation. B: Absence of ICA diastolic flow (red arrow) indicates more distal occlusion or severe stenosis

of the extracranial or intracranial ICA. C: Normal external carotid artery waveform. The external carotid artery has low diastolic

flow reflecting the high resistance circulation of the scalp and face. D: Normal common carotid artery waveform. The common

carotid artery waveform has diastolic flow well above the baseline as the majority of flow is directed to the low resistance internal

carotid artery circulation.

2488

Figure 87-3. In a patient with internal carotid occlusion common carotid artery (CCA) diastolic flow approaches 0 (arrows) as CCA

flow is now routed to the high resistance external carotid circulation.

The CCA contralateral to an ICA stenosis or occlusion may demonstrate an increased flow velocity

with particular elevation of the velocity at the end of diastole; the so-called end-diastolic velocity

(EDV). These changes represent a compensatory increase in blood flow volume in the nonobstructed

ICA in response to reduced contribution to cerebral perfusion from the opposite high-grade stenotic or

occluded ICA. This compensatory hemodynamic change can be substantial and flow velocities may be

artificially elevated on the side with compensatory high-volume flow giving a false impression of the

presence of a stenosis or a higher degree of stenosis than is actually present.2

In the presence of a significant stenosis at the origin of a CCA or the inominate artery, the ipsilateral

CCA waveform may be dampened, with low overall peak systolic velocities (PSVs) and EDVs and a

slower rise to peak systole when compared with the contralateral CCA waveform and potentially

reverse flow in the ICA (Fig. 87-4A,B). The CCA flow changes seen with proximal stenosis are also

important in the diagnosis of ICA stenosis because the overall reduction in flow velocity may artificially

lower velocities in an ipsilateral ICA stenosis, leading to underestimation of the severity of ICA stenosis.

In some cases of proximal CCA or innominate artery stenosis, the stenotic lesion may not be accessible

to direct insonation by the duplex scanner. In such cases, a stenotic lesion can be suspected as the

ipsilateral CCA waveform will often exhibit poststenotic turbulence low in the neck, reflecting disturbed

flow distal to the more proximal stenosis.

ICA Waveforms

As noted above, the normal ICA spectral waveform is indicative of high flow in a low-resistant

circulation. The systolic upstroke is rapid, PSV is <125 cm/s, and flow is maintained throughout

diastole (Fig. 87-2A).3 In the absence of plaque there is generally a clear spectral window under the

outline of the spectral waveform as there is little turbulent flow. The presence of color shifts, indicating

high-velocity flow, and color mosaics, indicating poststenotic turbulence, aids in selecting potential

areas for examination with the pulsed Doppler.

Hemodynamic quantification of the severity of ICA stenosis is primarily achieved by analysis of

Doppler spectral waveforms and measurements of peak systolic and EDV or comparison of PSVs in the

ICA to those in the CCA just proximal to the carotid bifurcation, ICA/CCA ratio. As a stenosis develops,

the PSV first becomes elevated. PSV is the primary measure of stenosis severity. EDV lags behind,

relatively speaking, as stenosis severity progresses but rises rapidly as the stenosis becomes severe

(diameter reductions of ≥60%). Thus, elevation of EDV is a good marker for high-grade stenosis (Fig.

87-5).4 The ICA/CCA ratio is also a very important measure of stenosis severity.5 Because it is a ratio, it

compensates for abnormally high- and low-flow states that may skew the PSV and EDV upward or

downward.

To accurately measure flow velocity, the sample volume must be placed within the area of greatest

stenosis. Color flow imaging has demonstrated that the orientation of the stenotic jet within a stenosis is

frequently not along the longitudinal axis of the vessel. This finding has resulted in controversy with

2489