439

Downloaded from www.ajronline.org by 177.128.8.26 on 02/09/24 from IP address 177.128.8.26. Copyright ARRS. For personal use only; all rights reserved

Review Article

The Doppler Signal: Where Does It Come From and What

Does It Mean?
T. R. Nelson1 and D. H. Pretorius

Sonography has become one of the most common tech- flow Doppler equipment; and, finally, to identify limitations and
niques for noninvasive evaluation of the body. It has taken on artifacts introduced by Doppler equipment.
a major role in obstetrics largely because of its minimal
bioeffects, flexibility, and low cost. Sonography also offers
Physics of Doppler Sonography
several distinct diagnostic tools in one instrument, including
imaging and Doppler, enabling both anatomic and dynamic The Doppler effect (named after C. J. Doppler, 1803-1853)
information to be obtained in real time. is really quite simple, and intuition is an excellent guide in
Recent developments in instrumentation, largely an out- understanding its relation to sonography. Several excellent
growth of the advances in miniaturization of electronics, now texts discuss physical acoustics related to sonography, so
provide information of a physiologic as well as an anatomic only a brief review of the basic concepts is given here [1-8].
nature. Combining anatomic and physiologic data adds an An ultrasound acoustic wave is a longitudinal compres-
additional dimension. However, more complex information sional wave consisting of a series of compressions and rare-
also places additional demands on the diagnostician, because factions (Fig. 1). Compressions represent areas of increased
more extensive knowledge and understanding are now re- pressure, and rarefactions represent areas ofdecreased pres-
quired about both the anatomy and physiology of the organ sure with respect to the ambient pressure. The characteristics
being imaged. of acoustic waves that are important for understanding the
More complex instrumentation also places additional de- Doppler effect are propagation velocity, frequency, reflection,
mands on the clinician to appreciate the operational principles and scattering. The pressure wave moves through the me-
and inherent limitations imposed on any study. Although the dium at a characteristic propagation velocity for each tissue
concepts underlying Doppler sonography are not complex, and depends on the elastic modulus and density of the tissue
mastery of the technique requires extra training. In addition, medium.
a significant commitment to learning the subtleties of clinical The elastic modulus is a tissue characteristic that relates
applications is necessary. The quality of clinical studies is how well the molecules and cells in a small tissue volume are
operator-dependent to an even greater degree than in sono- connected or tied together. The density is a measure of how
graphic imaging, thus making adequate training essential. The many atoms are present in that small tissue volume. Because
purpose of this article is to review the basic concepts con- the elasticity and density are tissue characteristics, it follows
cerning the physics of Doppler sonography; to set forth the that each tissue also has a characteristic velocity. The char-
rationale for Doppler investigation of basic physiologic proc- acteristic velocity of sound in tissue is nominally 1540 rn/sec.
esses as related to blood flow; to discuss instrumentation Actually, each tissue has a characteristic velocity, but 1540
including continuous-wave, pulsed-wave, duplex, and color- rn/sec represents a good average value. The sound wave

Received January 11, 1988; accepted after revision May 4, 1988.

‘Both authors: Department of Radiology, M-010, University of California, San Diego, La Jolla,CA 92093. Address reprint requests to T. A. Nelson.
AJR 151:439-447, September 1988 0361 -803x/88/1 51 3-0439 © American Roentgen Ray Society
 440 NELSON AND PRETORIUS AJR:151, September 1988

Pulse Length Axial Resolution Frequency Bandwidth

A fliL /‘TN
Downloaded from www.ajronline.org by 177.128.8.26 on 02/09/24 from IP address 177.128.8.26. Copyright ARRS. For personal use only; all rights reserved

0 B-IF--
I-

C,)
C,)
0
I-
0
c1U II
Time Distance Frequency
Time -

Fig. 1.-Diagram of acoustic wave. Fig. 2.-Relationship between pulse length, axial resolution, and fre-
Top, Vibrations are in same direction as propagation of wave. X = quency bandwidth for three different pulse lengths. Short pulses (A) have
wavelength. best axial resolution, resolving all four surfaces of the vessel. However,
Bottom, Time-varying changes in localized pressure due to passing there Is also a very broad band of frequencies centered at the nominal
wave. High pressure produces areas of compression and low pressure transducer frequency. intermediate (B) and long (C) pulse lengths have
produces areas of rarefaction. decreased axial resolution but a narrowed frequency bandwidth. Because
both narrow-frequency bandwidth and good axial resolution are desirable
characteristics in Doppler instruments, all equipment contains a trade-off
between these features.

also has a characteristic frequency (number of pressure peaks As long as the acoustic characteristics of the tissues in the
per second) and wavelength (distance between pressure sound field are constant, the wave will continue to propagate
peaks), both of which depend on the transducer design. The away from the transducer. However, when a tissue with
relationship between the frequency, wavelength, and velocity different acoustic characteristics is encountered, some of the
is: sound energy will be reflected back to the transducer. The
relative amount of energy reflected back depends on the
velocity = frequency x wavelength (1) differences between the two tissues. The acoustic properties
Because the velocity is constant, the wavelength must de- of a tissue are expressed as its acoustic impedance:
crease as the frequency of the sound wave (transducer) impedance = velocity x density
increases. According to basic physics, as the wavelength
decreases, the resolution increases; thus it is desirable to The amount and type of reflection depend on the relative
image with the highest possible frequency. impedance and the relationship between the size of the
Doppler information may be obtained with either continuous interface and the wavelength (A). That is, reflection at an
or pulsed ultrasound waves, whereas imaging uses only short interface with a large area interface, such as between the
pulses of ultrasound. The major difference between pulsed liver and diaphragm (where X is much less than the object
and continuous waves is the amount of time that the beam is size), will be specular, like a mirror. Intermediate-size inter-
turned on. The time for which the beam is on is important for faces (where A is about the same size as the object) will
tissues in respect to energy transfer and power deposition. diffract the beam. A small object, such as an RBC (where A
Pulse length also affects two important characteristics of the is much larger than the object size), will scatter the acoustic
acoustic wave. First, the distribution of frequencies present wave in all directions. The amount of scattering depends on
in the beam varies with pulse length: thefrequency distribution the scatterer’s radius raised to the sixth power and the
broadens as the pulse gets shorter [9]. Second, the axial frequency raised to the fourth power, suggesting a very
resolution of the imaging system depends on the pulse length: strong size dependence. If there is no interface, no energy
the shorter the pulse length, the better the axial resolution. will be reflected. On the other hand, if there is a very large
Short pulses have the broadest frequency distribution but the acoustic impedance difference (e.g., tissue-air), then essen-
best axial resolution. Continuous-wave beams have the nar- tially all the energy will be reflected. Most medical sonographic
rowest frequency distribution; however, they have no axial signals originate somewhere between these two extremes.
resolution. Thus, there is a trade-off between frequency Although images are dependent on all three types of interac-
broadening and axial resolution that is important in the design tions, the Doppler signals of interest are predominantly the
of clinical equipment (Fig. 2). result of scattering from RBCs.
Ultrasound waves will propagate until all of their energy is
dissipated in the tissue. In general, the amount of attenuation
The Doppler Effect
increases with distance from the transducer. The relative
amount of attenuation per unit distance (dB/cm/MHz) also Movement of the reflector with respect to the sound source
increases with the frequency. For this reason, superficial (e.g., movement due to flowing blood cells) results in a change
imaging is able to use higher frequencies than is deep imaging. in wavelength (Fig. 3). Since the velocity of sound is constant,
 AJA:151, September 1988 DOPPLER SIGNAL 441

the change in the wavelength corresponds to a change in for investigating the vessel of interest. The vessel (e.g., portal
frequency. The magnitude of the change in frequency de- vein, superior mesenteric artery, umbilical vein) frequently is
pends on (1) the wave velocity in the medium (C), (2) the parallel to the skin surface and hence near a right angle to
relative velocities of the reflector and the source (V), and (3) the beam axis. Second, the signal-to-noise ratio ultimately
the frequency of the source (Fo). The frequency change, limits the sensitivity of the system, and as the angle increases,
therefore, is proportional to the relative velocity ofthe reflector there is a decreased frequency shift. A higher frequency
and the propagation velocity of the wave in the medium. transducer may be used that will produce a greater frequency
Downloaded from www.ajronline.org by 177.128.8.26 on 02/09/24 from IP address 177.128.8.26. Copyright ARRS. For personal use only; all rights reserved

Movement of the reflector with respect to the sound source shift, partially overcoming the signal-to-noise limitations.
results in a change in frequency at the receiver that is actually Although many types of instrumentation present Doppler
doubled. The frequency change is doubled because both (1) information as the frequency shift (in kilohertz), comparison
the relative velocity between the source and the reflector and between measurements made on different instruments is
(2) the relative velocity between the reflector and the receiver difficult because the transducer frequency and angle may be
contribute to the change in frequency. The detected frequency different. Doppler shifts are only comparable when the fre-
change (SF) is: quency and angle are given. Comparison of measurements
from repeated studies or different laboratories is difficult, if
F = 2VF0/C (3)
not impossible, to make when different transducers are used.
This equation shows that the greater the transducer fre- More recent equipment displays the velocity directly and
quency (F0), the greater the frequency difference (SF) (and includes the angular correction in the calculation:
indirectly sensitivity) for a given object velocity.
V = FC/(2F0 cos 0) (4)
When the motion of the reflector and the sound source are
not parallel, it becomes necessary to correct for the angular
To summarize briefly, the Doppler shift usually encountered
differences. The frequency shift is measured only for that
clinically covers a velocity range of 1 0-1 00 cm/sec and rep-
component of motion occurring along the axis of the trans- resents a frequency shift range of 1 00 Hz to 1 1 kHz depend-
ducer beam. As a result, there will be a reduction in the
ing on the transducer frequency. Velocities encountered with
magnitude of the frequency shift, and thus in measured
stenoses may be even higher, approaching upward of 500-
velocity, that is solely due to the angle.
600 cm/sec. These frequency shifts are in the audible range
Velocity measurement errors due to angular variations have
and may be heard as well as seen during the study. The
two consequences that are important. First, absolute velocity
nominal velocity of ultrasound in tissue is approximately 1540
measurements require that a correction be made for the angle
m/sec. Clinically used transducers cover the range of 1-10
between the vessel and the axis of the ultrasound beam. MHz [10, 11].
Correcting for the angle also increases any uncertainty about
the velocity measurement. Thus, the larger the angle is, the
larger the potential error is. Large angles are encountered Physiologic Aspects of Blood Flow
often when deep structures are being evaluated. As a result,
there is reduced flexibility in selecting a more optimal angle Doppler velocity information is an important indication of
underlying physiologic processes. The relationship between
blood pressure, flow, and velocity is important to our under-
standing of the Doppler information obtained by sonography.
Blood is a moving liquid consisting of cells and plasma fluid.
fin nnn np )lii. A
The viscosity of the blood is directly related to the relative
(U percentage of RBCs in the blood (hematocrit). Because most
(((( of the resistance to blood flow in the circulatory system occurs
F<F f<0
R S in the small vessels and capillaries, the viscosity is important
in determining the amount of work necessary to push the
LIm ))iF=F )yyyy
F0 .
blood through the system. The blood viscosity also increases
dramatically as the velocity of blood flow decreases.
LR J((( (((((( (((((( Two factors affect blood flow through a vessel: (1) the
R S
pressure differential between the ends of the vessel and (2)
the resistance to flow imposed by the vessel. The relationship
between flow, pressure differential, and resistance in a vessel
tiiii:ii:iii))i) )i)I) )I\(( iki #{149}---
is expressed as:
Lfl ((((U ((((U ((((U flow = pressure differential/resistance (5)
F >F F>0
R S

Blood flow is the total amount of blood moving (at all

Fig. 3.-Doppler effect showing changes in wavelength (X) resulting
from moving (A and C) and stationary (B) reflectors (RBCs). Frequency
velocities) past a certain point. Therefore, if we measure the
shift (tF) depends on both the relative velocities (1) between the source entire velocity distribution, we can estimate the flow (in ml/
(5) and the reflector (R) and (2) between the reflector and the receiver. sec) through the vessel. The resistance to flow is dependent
Movement toward the transducer (C) produces a positive frequency shift,
while movement away from the transducer (A) produces a negative fre- on the radius of the vessel, the length of the vessel, and the
quency shift. viscosity of the blood and is given by [12]:
 442 NELSON AND PRETORIUS AJR:151, September 1988

viscosity x length 6 does because blood in contact with the vessel wall is slower
resistance . 4
radius ( ) than blood in the center of the stream. Turbulent flow has a
broad velocity distribution, with velocity components ranging
The radius is the critical factor determining the resistance from very slow to very fast jets. In addition, eddies and
because resistance is inversely proportional to the fourth vortices give rise to velocity components in the opposite
power of the radius. This means that small changes in vessel direction. Reverse flow is encountered frequently in turbulent
radius will produce dramatic changes in resistance and thus velocity patterns associated with stenotic obstructions. Re-
Downloaded from www.ajronline.org by 177.128.8.26 on 02/09/24 from IP address 177.128.8.26. Copyright ARRS. For personal use only; all rights reserved

flow. We also know that: verse flow also may be encountered at vessel bifurcations
flow (cm3/sec) =
where boundary layer separation occurs. The type of flow
mean velocity (cm/sec) x area (cm2) (7) depends on (1) the velocity of the blood, (2) the viscosity of
blood, (3) the radius of the vessel, and (4) the density of
In principle, sonography makes possible the estimation of blood. The relationship between these parameters is ex-
volume flow (in mI/sec) by careful measurement of the area pressed via the Reynolds’ number [12]:
(image) and the velocity distribution (Doppler trace) [1 3, 14].
Flow measurements, in turn, provide indirect information Reynolds’ number =
about the peripheral vascular resistance. (velocity x radius x density)/viscosity (8)
Because the circulation is a pulsatile system, the velocity
distributions and waveforms are constantly changing, albeit The Reynolds’ number is a measure of the tendency for
cyclically. The pressure pulses and resultant flow changes turbulence to occur. Turbulence will generally occur if the
are reflected in cyclic changes in the velocity curves. In Reynolds’ number exceeds (1) 2000 for large vessels, (2)
addition, each vessel has a characteristic signature that re- 1 000 for straight arteries, and (3) 200 for branches in the
flects its relative position and physiologic status in the system arterial tree. Thus, the threshold value of the Reynolds’ num-
[1 5]. Waveforms also are affected by the body’s cardiac ber for turbulence to occur depends on the geometry of the
output, gravity orientation, exercise, stress level, and diges- vessel and tends to decrease with increasing vascular corn-
tion. Differences between systole and diastole frequently plexity.
reflect important physiologic information as well and have The most obvious difference between plug or laminar and
been compared by using various indexes that will be explained turbulent flow is the relative distribution of velocities. As a
later. result, the Doppler spectrum will be much broader with tur-
The pulsatile nature of the circulation leads to several types bulent flow than with plug or laminar flow. Turbulence may
of flow and associated velocity distributions (e.g., plug, lami- occur at any point where a disruption to smooth flow might
nar, turbulent) (Fig. 4). Plug flow has a very narrow velocity be present, such as bifurcations, sharp bends, plaques in
distribution because all the blood is moving at essentially the vessels, or pathologic narrowing of vessels. Turbulence from
same velocity. Plug flow is found frequently in pulsatile sys- atherosclerotic narrowing of vessels and the associated irreg-
tems, particularly the larger vessels such as the aorta. Lami- ular luminal surfaces often will show a very broad velocity
nar flow has a broader velocity distribution than plug flow distribution as well as reverse flow resulting from eddy cur-
rents and vortices.

Instrumentation
Continuous-Wave Doppler
One of the most straightforward methods of obtaining
Doppler velocity information is with continuous-wave tech-
niques. Essentially, the transmitter is on continuously; a sep-
arate transducer receives the signal and compares it with the
Plug Laminar Turbulent transmitted reference. The frequency-difference signal may
be presented audibly as well as graphically. Continuous-wave
Doppler systems may measure a wide range of velocities
-
without limit. However, there is no position information and it
- is not possible to determine the exact source of the Doppler
signal because any moving object in the beam will produce a
I I vA0.fL1j __________ signal (Fig. 5).

Pulsed-Wave Doppler
Velocity Distribution The position insensitivity of continuous-wave Doppler is
overcome to a large extent by using pulsed-wave trans-
Fig. 4.-Plug, laminar, and turbulent flow. Plug flow consIsts of blood
moving at euentially a single velocity. LamInar flow contains components
ducers. A sound pulse is transmitted and the returning echo
at many velocities. Velocity (V) profile generally is parabolic, with slowest is detected by a single transducer assembly. The time delay
movement occurring at Interface between vessel and blood; fastest movIng between the pulse and the returning echo depends on the
blood Is In vessel center. Turbulent flow contains a broad spectrum of
velocity components, including components due to Jets and reverse flow
velocity of sound and directly relates to the depth of the
that form downstream from obstructions. reflector (Fig. 5). The important parameters are the duration
 AJR:151, September 1988 DOPPLER SIGNAL 443

dous amount of velocity information requires presenting the

data in a manner more readily assimilated by the observer.
+ MRMM A
Color-flow instrumentation extracts velocity information from
the returning echoes and adds the information to the conven-
\!VVV\&& B
tional two-dimensional image as color information. The color
and intensity represent the direction and magnitude of the
4
Downloaded from www.ajronline.org by 177.128.8.26 on 02/09/24 from IP address 177.128.8.26. Copyright ARRS. For personal use only; all rights reserved

velocities present in the image (Figs. 7 and 8). The diagnos-

M. A
tician may also select areas in the image for evaluation of
B
their temporal characteristics as is done with duplex Doppler
imaging. Color-flow sonographic imaging combines both an-
atomic and velocity data and displays the information to the
clinician in a format that can be easily comprehended. Detailed
Fig. 5.-Continuous-wave and puised-wave Doppler signals. quantitative analysis still requires range-gate Doppler mea-
Lift, Transducerinterrogatestwovesseis(A and B)wlth different depths
surement. Low-velocity amplitudes may not be visualized
and different directions of blood flow.
Right, Upper tracing is representative of information obtained from because of display thresholds and averaging effects. Routine
continuous-waveequipment and illustrates signals detected from both studies may be performed more rapidly because complex
vessils A and B sImultaneously. Resulting trace has both positive (toward
transducer) and negative (away from transducer) components from the
vascular structures may be identified visually.
two vessels. Lower tracing is from pulsed-wave Doppler equipment and
makes ft possible to select either vessel A (solldlln.s) or vessel B (broken
lhs) for evaluation. Factors Affecting Velocity Waveforms and Distributions

Angular Dependence
of the ultrasound pulse (pulse length) and the relative amount Accurate velocity measurements require that a correction
of time that the beam is turned on (duty cycle). If motion is be made for the angle between the axis of the beam and the
present, there is also a frequency shift in the returning pulse. vessel because vessels are rarely imaged along the axis. The
Parameters including pulse length and duty cycle place an anatomy of the vessel needs to be understood dearly in order
upper limit on the maximum velocity that may be measured. to produce an accurate correction. Tortuous vessels con-
In addition, because the delay is related to the depth of the stantly change direction. As a result, accurate determination
echo, it is possible to measure selectively the velocity at of the proper correction angle for velocity measurements is
specific locations in the beam. However, the precise source extremely difficult. The angle between the vessel and the
of the signal is difficult to determine because a picture of the ultrasound-beam axis should be as small as possible (<60#{176})
subsurface anatomy is not shown. in order to produce signals with large signal-to-noise ratios
(Fig. 9). The operator also must select the angular correction
Duplex Doppler/Imaging for the velocity measurement, or an erroneous velocity mea-
surement may be produced by incorrect selection of the
Duplex imaging combines pulsed-wave Doppler with two- angular compensation. In addition, some equipment will report
dimensional real-time imaging; this allows for precise localiza- a velocity whether or not the angular correction has been
tion of the Doppler sample. Duplex imaging is possible with a made, leading to potentially significant misreporting of velocity
single machine that produces both image and velocity infor- data. Figure 10 shows the effect of incorrect angular corn-
mation (Fig. 6). However, because the optimal requirements pensation on measured velocity values for samples taken
for imaging and velocity measurement are slightly different, within the same vessel. The vessel geometry proximal and
many systems use two transducers located in one probe for distal to the sampling site must be determined in order to
duplex imaging. The two data sets are combined on the real- make a proper angular correction for accurate velocity mea-
time display and indicate the location of the Doppler measure- surement.
ment. An indication of the sample size is also provided on the
display. The change in blood velocity with time is shown as a
moving trace similar to an ECG trace. From this display, The Effect of Sample Volume Size
information can be obtained about the peak velocity as well The sample volume size, waveform, velocity distribution,
as about the velocity distribution for any point in the cardiac and volume flow measurements all depend on the gate size.
cycle. Duplex Doppler imaging offers many advantages over The effect of changing the gate size and location within the
either continuous- or pulsed-wave Doppler, and it has been vessel is shown in Figure 1 1 When the gate is located in the
.

this technology that has allowed Doppler to become so clini- vessel center, the peak velocity measurement will be the
cally useful in evaluating carotid vessels [1 6-20], fetal and same for all gate sizes. However, each larger gate includes a
uterine vessels [21 -25], and abdominal vessels [1 0, 15, greater range of velocity distributions, and thus the distribu-
26-28]. tion “fills in” under the wpeak value. If the vessel wall also is
moving, a strong low-velocity signal is obtained that may
obscure the blood signal. Selective filters (wall filters) may be
Color-Row Doppler Imaging
used to remove this unwanted information.
Color-flow Doppler imaging makes available to the diagnos- The intensity of the Doppler signal is determined by the
tician velocity information for the entire image. The tremen- number of scatterers in the gate volume. Careful measure-
 444 NELSON AND PRETORIUS AJR:151, September 1988

Fig. 6.-Images from duplex instru-

ment. Doppler velocity display is shown
in lower panels with imaged vessel in
upper panels. Peak velocIty (0.85 m/
eec) is the same for all images.
Left, image Is typical trace from in-
ternal carotid artery at standard set-
tings.
Center, increased receiver gain Is
Downloaded from www.ajronline.org by 177.128.8.26 on 02/09/24 from IP address 177.128.8.26. Copyright ARRS. For personal use only; all rights reserved

shown.
Right, Decreased receiver gain is
shown.
inappropriate gain settings could
lead to misInterpretation of spectral
broadening

Fig. 7.-Images obtained from color-flow instrument (red represents blood moving toward the transducer blue represents blood moving away) for
velocity range settings of 0.09 m/sec (A), 0.48 rn/sec (B), and 1.6 m/sec (C). improper velocity range selection leads to ailasing In A or limited dynamic
range in C. Ailasing can be a significant problem In color-flow imaging If it is not recognized because ft will appear sImilar to reverse flow.

Fig. 8.-Color-flow images displayIng the effect of different reference angles for the Doppler measurement. The Doppler angie is parallel to the angled
box in all three images. Thus, without moving the transducer, we can see velocity components relative to the uftrasound-beam axis for three different axis
directions (A, B, C). in each image, red represents velocity components parallel to the beam axis and toward the transducer, whereas blue represents
velocity components away from the transducer. All three images are from the same vessel location, and blood is movIng from left to right continuously.
The direction of the ultrasound-beam axis is the only difference between images.

ment of the signal intensity for each velocity in the vessel The selection of gate size is governed by the size of the
permits calculation of total flow. However, extreme care is vessel and the signal-to-noise characteristics of the signal.
necessary to convert the velocity information to flow infor- Generally, it is desirable to measure the peak velocity, and
mation because all velocity components must be included. thus a smaller gate is used than would normally be necessary
 AJR:151, September 1988 DOPPLER SIGNAL 445
Downloaded from www.ajronline.org by 177.128.8.26 on 02/09/24 from IP address 177.128.8.26. Copyright ARRS. For personal use only; all rights reserved

Fig. 9.-Velocity traces showing difference in signal and data for dif-
ferent angles between acoustic axis and vessel axis for the same vessel
and gate size.
Left, image is properly set up (the dots represent Doppler beam axis)
and shows a normal internal carotid trace.
Right, image displays incorrect angle for obtaining Doppler Information.
Even though the gate is in the same location, the angle is very near to 900
and requires a computational correction that is very large. Small differ-
ences In signal due to poor signal-to-noise ratio are magnified, and trace
yIelds Inaccurate data with artificially high-velocity components and sig-
nificant distortion.

Fig. 11.-Effect of changing gate size and position on observed and

measured velocIty distribution.
upp.rwt, image uses small, centered gate and produces clean Doppler
trace (although ft Is not suitable for flow measurements).
upper rIght, Image uses large gate encompassing entire vessel. Re-
suiting Doppler trace, assuming amplitude quantification is possible, could
be used to compute flow. Peak velocity values are the same in both
images.
Lower left, Image has moved small gate to position near vessel wall.
Peak measured velocity is sllghtiy decreased; in addition, lower velocity
components are also detected compared with upper left image.
Lower right, Image has placed gate on vessel wail. Waveform and
magnitude of velocity distribution are greatly altered. It is essential that
gate be centered in vessel axially to produce optimal velocity (and flow)
measurements.

Fig. 10.-Doppler traces for the same geometry as in Fig. 9 with an

incorrect angular correction selected (small line through gate shouid be
set parallel to vessel axis). Although waveform is the same in both left and the vessel depth, the transmitter frequency, and the angle to
right images, computed value for velocity is different, reflecting compute- flow. Aliasing results in the display of erroneous velocity
tional error even though signal-to-noIse ratio is good.
information. Although errors may be obvious, complex yes-
sels or velocity may give misleading data because of this type
to obtain valid flow data. However, useful information is of artifact (Fig. 13). Aliasing is not a problem with continuous-
contained in the velocity distribution displayed on the image. wave Doppler instrumentation but only with instruments that
The intensity of the velocity display provides an indication of use pulsed, sampled techniques, including color flow. Aliasing
the relative amount of blood moving at a particular velocity. is not difficult to recognize and may be corrected by increasing
Also, vessels that contain plaques or severe atherosclerotic the velocity range limit of the instrument, changing to a lower
disease no longer have smooth laminar flow and have a transmitterfrequency, or increasing the angle toflow up to 60#{176}.
broader velocity distribution including higher as well as lower
and reverse velocities (Fig. 12). The relative increase in mid-
Indexes
velocity signal from the turbulent flow will be readily detect-
able. The distribution of velocity information is an important Doppler indexes have been used to obtain information
diagnostic parameter that may be used to evaluate both flow involving blood flow and vascular impedance that cannot be
and pathophysiology [29]. obtained from absolute velocity information alone. Several
indexes have been developed to evaluate and compare Dop-
Allasing pIer waveforms. The indexes depend on ratios involving the
Aliasing occurs when the velocity of the blood exceeds, an systolic peak velocity (A), the end-diastolic peak velocity (B),
upper measurement limit, defined by the pulse repetition rate, and the mean velocity throughout one cycle (mean). One
 446 NELSON AND PRETORIUS AJR:151, September 1988

Fig. 12.-Doppler tracings for tur-

bulent flow occurring at bifurcation with
three different gate positions and
sizes. Waveforms are greatly affected
by small changes In location of gate. in
addition, there is a generai filling in
under-the-peak values typical of the
broad range of velocities found In tur-
bulent sites.
Downloaded from www.ajronline.org by 177.128.8.26 on 02/09/24 from IP address 177.128.8.26. Copyright ARRS. For personal use only; all rights reserved

Fig. 13.-Dopplertraces showing al-

lasing of waveform due to improper
selection of velocIty range.
Left waveform, Normal trace.
Center and right, images show var-
bus amounts of aliaslng.
INVERTED Indicates that trace has
been reversed. While obvious, such
changes in “standard” display format
can produce Interpretational errors un-
less fully noted.

The A/B ratio has been used primarily in obstetric sonog-

A A
- index- (A-B) raphy for evaluating waveforms from the umbilical cord and
>‘
uteroplacental arteries, where it is difficult to determine vessel
0 Resistive index
0
angle or size [21]. In addition, many investigators believe that
‘I) Puisatility Index- Mean
measurement of placental impedance is more valuable than
> flow measurements in evaluating for intrauterine growth re-
tardation. The Pourcelot (or resistive) index is slightly more
End-Diastole
sensitive in differentiating abnormal waveforms because the
denominator never becomes zero [26]. This index has been
Time -
used primarily in evaluating renal transplants. The pulsatility
index is the most sensitive in differentiating abnormal wave-
Fig. 14.-Three commonly used clinical Indexes: systolic peak velocity
(A)lend-diastollc peak velocity (B) ratio, Pourcelot (resIstIve) index, and
forms because it takes into account the mean velocity [22].
utillty index. Depending on instrument sophistication, one or more of However, this calculation has not been made available on all
these indexes may be computed automatically. If not, the first two may be instruments, and therefore extra analysis is required. The
easily computed from hard-copy Images. Specific points for measurement
are indicated, as well as method of calculation. pulsatility index has been used in obstetric and abdominal
Doppler evaluations.
The factors affecting the intensity and signal-to-noise ratio
of the Doppler signal are still important when indexes are
used, however, and cannot be ignored. The locations of the
advantage of these indexes is that the angular correction and measurements are shown in Figure 14 for several different
vessel size cancel each other out and thus do not need to be definitions of pulsatility.
known. This is extremely important in evaluating small vessels
and vessels that move freely, such as those in the umbilical
cord. Signal-to-noise ratios still are important in terms of the Summary
accuracy of the measurement. The three primary indexes
used clinically are: Doppler sonographic measurement of blood velocity and
associated physiologic parameters is a powerful diagnostic
A/B ratio = A/B (9) technique. State-of-the-art instrumentation incorporates ye-
locity measurement with two-dimensional imaging capability;
resistive or Pourcelot index = (A - B)/A (10) it uses intensity and color coding to display complex physio-
logic and anatomic data to the observer in an easily under-
pulsatility index = (A - B)/mean (11) stood format. Although the concepts underlying Doppler so-
 AJR:151, September 1988 DOPPLER SIGNAL 447

nography are not complex, mastery of the technique requires 3. Rose JL, Goldberg BB. Basic physics in diagnostic ultrasound. New York:
Wiley, 1979:1-314
extra training and commitment. The principal features and
4. Powis AL, Powis WJ. A thinker’s guide to ultrasonic imaging. Baltimore:
clinical practicalities associated with Doppler sonography are Urban & Schwarzenberg, 1984:3-413
summarized in the following paragraphs. 5. Bushong Sc. Radiologic science for technologists. St. Louis: Mosby,
Continuous-wave Doppler is very sensitive to small vessels 1984:1-621
6. Kisslo J, Adams D, Mark DB, eds. Basic Doppler echocardiography. Clln
and has no upper velocity limit. In addition, the instrumentation
Diagn Ultrasound 1986;17: 1-208
Downloaded from www.ajronline.org by 177.128.8.26 on 02/09/24 from IP address 177.128.8.26. Copyright ARRS. For personal use only; all rights reserved

is not complex and produces relatively low acoustic power. 7. Walls PNT. Biomedical ultrasonics. New York: Academic Press, 1977:
A significant drawback to continuous-wave Doppler is that 1-635
there is no depth sensitivity, and thus complex structures or 8. Walls PNT. Utrasonics in medicine and biology. Phys Med Blol
multiple vessels can give conflicting information. 1977;22:629-669

Pulsed Doppler (including duplex and color-flow) instrumen- 9. Holland 5K, Orphandoudakis Sc, Jaffe cc. Frequency dependent atten-
uation effects in pulsed Doppler sonography: experimental results. IEEE
tation has the capability of depth resolution and a variable Trans Biomed Eng 1984;BME-31 :626-631
sample volume. Pulsed Doppler equipment is prone to aliasing 10. Taylor KJW, Bums PN. Doppler sonography: continuous and pulsed,
(false velocity indications) under some circumstances and also superficial and deep. Categorical course. Radiology 1984;153(P):225-226
11. creeimore SP, Graham MM, Jahn GE, Targett RC, Mcllroy MB. Compar-
produces higher peak power levels than does continuous-
icon of methods of recording and analysis of Doppler blood velocity signals
wave equipment. Duplex equipment is more complex and in normal subjects. Ultrasound Med Bid 1982;8:525-535
expensive than continuous-wave equipment because the two- 12. Guyton A. Physics of blood, blood flow and pressure: hemodynarncs. In:
dimensional and Doppler modes must be synchronized in Textbook of medicalphysiology. Philadelphia: Saunders, 1971:204-217
operation and display. Color-flow equipment is extremely 13. Gill RW. Pulsed Doppler with B-mode imaging for quantitative blood flow
measurement. Ultrasound Med Bid 1979;5:223-235
complex and expensive. Color flow provides information of a
14. Walter JP, McGahan JP, Lantz BM. Absolute flow measurements using
qualitative and limited quantitative value. Absolute measure- pulsed Doppler ultrasound. RadiOlOgy 1986;159:545-548
ment still requires range-gate measurements. 15. Taylor KJW, Bums PN, Woodcock JP, Walls PNT. Blood flow in deep
Technical and anatomic factors will affect the measured abdominal and pelvic vessels: ultrasonic pulsed-DOppler analysis. Radio!-
velocity profiles. Thus, it is important to fully appreciate the ogy 1985;154:487-493
16. Wetzner SM, Kiser LC, Berzreh JS. Duplex sonography imaging: vascular
anatomy of the vessel and the angle between the vessel and applications. Radiology 1984;150:507-514
the ultrasound beam when making quantitative measure- 17. Green FM, Beach K, Strandness DE, Fell G, Philips DJ. Computer based
ments. Measurements that evaluate the velocity waveform pattern recognition of carotid arterial disease using pulsed Doppler sonog-
and make use of ratios, such as the pulsatility index, eliminate raphy. Ultrasound Med Bid 1982;8:161-176
18. Jacobs NM, Grant EG, Schellinger D, Byrd MC, Richardson JD, Cohen
the need for angular corrections; however, artifacts due to
SL. Duplex carotid sonography: criteria for stenosis, accuracy and pitfalls.
unappreciated anatomic or wall characteristics may lead to Radiology 1985;154:385-391
incorrect information if all parameters are not fully understood. 19. Garth KE, Carrol BA, Sommer FG, Oppenheimer DA. Duplex sonography
Doppler sonographic measurements may be used to deter- scanning of the carotid arteries with velocity spectrum analysis. Radiology
1983;147:823-827
mine the presence of flow, determine the direction of flow,
20. Zweibel WJ, Crummy AB. Sources of error in Doppler diagnosis of carotid
identify time-varying velocity characteristics, and detect ye- occlusive disease. AiR 1981;137:1-12
locity disturbances. Because flow and velocity are related, it 21 . Trudinger BJ, Giles WB, Cook CM, BOmbardIeti J, Collins L. Fetal umbilical
is possible to estimate flow from velocity measurements with artery flow velocity waveforms and placental resistance: clinical signifi-
careful calibration and proper precautions. Velocity is related cance. Br J Obstet Gynaeco! 1985;92:23-30
22. Wladimiroff JW, Wijngaard J. Degani S. Noordam PM, vEyck J, Tonge HM.
to flow, which, in turn, is related to both pressure and vascular
Cerebral and umbilical arterial blood flow velocity waveforms in normal and
resistance. Doppler velocity distributions also may be used to growth-retarded pregnancies. Obstet Gyneco! 1987;69:705-709
deduce acceleration and deceleration characteristics of blood 23. Arduini D, Rizzo G, Romanini C, Mancuso S. Fetal blood flow velocity
flow. waveforms as predictors of growth retardation. Obstet Gyneco!
1987;70:7-10
24. Erskine ALA, Ritchie JWK. Quantitative measurement of fetal blood flow
using Doppler sonography. Br J Obstet Gynaecol 1985;92:600-604
25. Eik-Nes 5, Marsal K, Knstoffersen K. Methodology and basic problems
ACKNOWLEDGMENTS
related to blood flow studles in the human fetus. Ultrasound Med Bid
1984;10:329-337
We thank John Forsythe, Division of Sonography, Veterans Ad-
26. Rifkin MD, Needleman L, Pasto ME, et al. Evaluation of renal transplant
ministration Medical Center, San Diego, for his assistance and ex- resection by duplex Doppler examination: value of the resistive index. AiR
pertise in obtaining the images used in this article. 1987;148:759-762
27. Taylor KJW, Morse 5S, Rigsby SM, Bia M, Schiff M. vascular complica-
boris in renal albografts: detection with duplex Doppler ultrasound. Radio!-
ogy 1987;162:31-38
REFERENCES 28. Borland LL, Lawson TL, Adams MB, Melrose BL, Foley WD. Evaluation of
renal transplants with pulsed Doppler duplex sonography. J Ultrasound
1. Hagen-Ansert SL Textbook of diagnostic ultrasonography. St. Louis: Med 1982;1 :215-222
Mosby, 1983:1-617 29. Knox RA, Phillips DJ, Breslau PJ, Lawrence A, Primozich J, Strandness
2. Woodcock JP, Skidmore R. Principles and applications of Doppler ultra- DE. Empirical findings relating sample volume size to diagnostic accuracy
sound. Clin Diagn Ultrasound 1980;5: 166-1 85 in pulsed Doppler cerebrovascular studies. JCU 1982;10:227-232