Chapter

6 Ultrasound
imaging
submarines, and in
Introduc seismology to
tion locate gas fields.
The basic
Ultrasound
principle of
imaging has been
ultrasound
used in clinical
imaging is simple.
practice for more
A propagating
than half a
wave partially
century. It is
reflects at the
noninvasive,
interface between
relatively
different tissues. If
inexpensive,
these reflections
portable, and has
are measured as a
an excellent
function of time,
temporal
information is
resolution.
obtained on the
Imaging by means
position of the
of acoustic waves
tissue if the
is not restricted to
velocity of the
medical imaging.
wave in the
It is used in
medium is known.
several other
However, besides
applications such
reflection, other
as in the field of
phenomena such
nondestructive
as diffraction,
testing of
refraction,
materials to check
attenuation,
for microscopic
dispersion, and
cracks in, for
scattering appear
example, airplane
when ultrasound
wings or bridges,
propagates through
in sound
matter. All these
navigation ranging
effects are
(SONAR) to
discussed below.
locate fish, in the
Ultrasound
study of the
imaging is used
seabed or to detect
 Chapter 6: Ultrasound imaging

not only to Essential to these

visualize developments
morphology or were the
anatomy but also publication of The
to visualize Theory of Sound
function by means by Lord Rayleigh
of blood and in 1877 and the
myocardial discovery of the
velocities. The piezoelectric effect
principle of by Pierre Curie in
velocity imaging 1880, which
was originally enabled easy
based on the generation and
Doppler effect and detection of
is therefore often ultrasonic waves.
referred to as
The first use of
Doppler imaging.
ultrasound as a
A well-known
diagnostic tool
example of the
dates back to 1942
Dopplereffectisthe
when two Austrian
suddenpitchchange
brothers used
ofawhistling train
when passing a transmission of
static observer. ultrasound through
Based on the the brain to locate
observed pitch tumors. In 1949,
change, the the first pulse-echo
velocity of the system was
train can be described, and
calculated. during the 1950s
Historically, 2D gray scale
the first practical images were
realization of produced. The first
ultrasound publication on
imaging was born applications of the
during World War Doppler technique
I in the quest for appeared in 1956.
detecting The first 2D gray
submarines. scale image was
Relatively soon produced in real
these attempts time in 1965 by a
were followed by scanner developed
echographic by Siemens. A
techniques adapted major step forward
to industrial was the
applications for introduction in
nondestructive 1968 of electronic
testing of metals. beam steering
12
9
 Chapter 6: Ultrasound imaging

using phased-array motion, as opposed to

technology. Since transverse waves,
the mid-1970s, such as waves on the
electronic scanners sea, for which this
have been displacement is
available from perpendicular to the
many companies. direction of
Image quality propagation. For
steadily improved compression waves,
during the 1980s, regions of high and
with substantial low particle density
enhancements are generated by the
since the mid- local displacement of
1990s. the particles. This is
illustrated in Figure
6.1. Compression
regions and
rarefaction regions
Physic correspond to high
s of and low pressure
areas, respectively.
acous Wave propagation
is possible thanks to
tic both the elasticity and
the inertia of the
wave medium: elasticity
s counteractsalocalcom
pressionfollowedbyare
What turnto equilibrium.
However, because of
are inertia, this return will
be too large, resulting
ultras in a local rarefaction,
onic which elasticity
counteracts again.
waves After a few iterations,
depending on the
? characteristics of the
Ultrasonic waves are medium and of the
progressive initial compression,
longitudinal equilibrium is reached
compression waves. because each iteration
For longitudinal is accompanied by
waves the damping. As
displacement of the
particles in the
medium is parallel to
the direction of wave
13
0
 Chapter 6: Ultrasound imaging

Figure6.1 Schematically, a longitudinal

wave can be represented by particles
connected by massless springs that are
displaced from their equilibrium
position. (From T.G. Leighton, The
Acoustic Bubble, Academic Press, 1994.
Reprinted with permission of Academic
Press, Inc.)

a consequence of generated and

these phenomena, the detectedbyapiezoelect
compression wave riccrystal.
propagates. Thesecrystalsdeform
The word under the influence of
“ultrasonic” relates to an electric field and,
the wave frequencies. vice versa, induce an
Sound in general is electric field over the
divided into three crystal upon
ranges: subsonic, deformation. As a
sonic and ultrasonic. consequence, when an
A sound wave is said alternating voltage is
to be sonic if its applied over the
frequency is within crystal, a compression
the audible spectrum wave with the same
of the human ear, frequency is
which ranges from 20 generated. A device
to 20000 Hz (20 kHz). converting one form
The frequency of of energy into another
subsonic waves is less form (in this case
than 20 Hz and that of electric to mechanical
ultrasonic waves is energy) is called a
higher than20kHz. transducer.
Frequenciesusedin(me
dical)ultrasound
imaging are about Wave
100–1000 times
higher than those propagation
detectable by humans.
in
homogeneou
Generation of
s media
ultrasonic This paragraph briefly
discusses the physical
waves phenomena observed
Usually ultrasonic during wave
waves are both propagation through
13
1
 Chapter 6: Ultrasound imaging

any homogeneous Linear wave equation

medium. It is Assuming small-
characterized by its amplitude waves
specific acoustic (small acoustic
impedance Z. As with pressures p)
any impedance in traveling through a
physics, this is the nonviscous and
ratio of the driving acoustically
force (acoustic homogeneous
pressure p) to the medium, the linear
particle velocity wave equation
response (v), i.e., p/v. holds
For plane, progressing
waves it can be shown
that ∇2p −

Z
1
(6.1) 2
∂
∂
where ρ is the mass 2 p
t
density and c the
acoustic wave 2 =
velocity in the 0,
medium. Table 6.1
(6.
illustrates that c and
consequently also Z 2)
c
typically increase with
ρ. where ∇2 is the
Table6.1 Values of
Laplacian. The
the acoustic wave
velocity c and
velocity c of sound
acoustic in soft tissue is
impedance Z of very close to that
some substances in water and is
Substance c approximately
(m/s)1540 m/s. In air
Air (25◦ C) 346 the sound velocity
Fat 1450is approximately
Water (25◦ C) 1493
300 m/s and in
bone
Soft tissue 1540
approximately
Liver 1550
4000 m/s.
Blood (37◦ C) 1570
Equation (6.2)
Bone 4000
is the basic
Aluminum 6320differential
equation for the
mathematical
13
2
 Chapter 6: Ultrasound imaging

description of superposition of a
wave propagation. left and right
A general solution progressive wave
of the linear wave with velocity c.
equation in one The shape of
dimension is p(x, t) the waveform is
= A1f1(x − ct) + irrelevant for the
A2f2(x + ct), (6.3) wave equation. In
other words, any
waveform that can
where f1(x) and
be generated can
f2(x) are arbitrary
propagate through
functions of x that
matter. A simple
are twice
example is the
differentiable. This
following
can easily be
sinusoidal plane
verified by
wave p(x, t) = p0
substituting Eq.
(6.3) into the wave
equation Eq. (6.2). sin
This solution is
illustrated in
Figure 6.2 for f1(x)
= f2(x) = f (x) and A1 , (6.4)
= A2. It represents
the where x is the
direction of the
p(x,t)
acoustic
2f (x) wave
propagation, λ the
wavelength, and T x

the period (c =
f (x + c0t ) f (x – c0t )
λ/T). In practice,
the shape of the
propagating wave
is defined by the
t characteristics of
the transducer.
Figure6.2 Schematic
representation of a Acoustic intensity
solution of the linear and loudness
wave equation. This
solution represents the The acoustic
superposition of a left intensity I (in units
and right propagating
wave. W/m2) of a wave
is the average
energy per time
unit through a unit
area perpendicular

13
3
 Chapter 6: Ultrasound imaging

to the propagation = I
direction. Hence, = 12
W/m2. (6.7)
T L 10 log10 I0 ,
1 with I0 10
I= p(
(6.5)
Using Eq. (6.6) L can
T also be written as
0
p
For the plane
progressive wave L = 20 log10 p 0 ,
of Eq. (6.4), the with p0 = 20 µPa.
acoustic intensity (6.8)
is
T I0 is the threshold
at 1000 Hz of human
= 1 hearing. Hence,
(t) dt absolute silence for
I p humans corresponds
Z to 0 dB. Increasing the
intensity by a factor of
ten corresponds to an
·
increase in sound level
of 10 dB and
T
0
approximately twice
T the perceived loudness
in
p 02
= · sin
− dt
= 0
Z 2Z
T humans. Doubling the
T intensity causes an
0 increase of 3 dB.
p2 The wave
. (6.6) frequency also has an
A tenfold increase effect on the perceived
in intensity sounds a loudness of sound. To
little more than twice compensate for this
as loud to the human effect, the phon scale,
ear. To express the shown in Figure 6.3,
sound level L (in is used. The phon is a
decibels (dB)) a unit of the perceived
logarithmic scale was loudness level for
therefore introduced pure, i.e., single-
frequency, tones. By
definition, the number
13
4
 Chapter 6: Ultrasound imaging

of phons equals the distance with respect

number of decibels at to the wavelength.
a frequency of 1 kHz. Figure 6.5 shows
In practice, such as in the simulated spatial
measuring equipment, distribution of the
the isophones are maximal pressure
often approximated by generated with a pulse
simplified weighting of 5 MHz by a circular
curves. For example, source with a diameter
the A-weighting, of 10 mm. When the
shown in Figure point of observation is
6.3(b), is an located far away from
approximation of the the source and on its
isophones in the range symmetry axis, the
20–40 phons.
A-weighted
measurements are
expressed in dB(A).

Interference
Interference between
waves is a well-
known phenomenon.
For an infinite (or a
very large) number of
coherentsources(i.e.,
sourceswiththesamefr
equency
andaconstantphaseshif
t)theresultingcomplexi
nterference pattern is
called diffraction. The
shape of this 3D
pattern is closely
related to the
geometry of the
acoustic source.
Figure 6.4 shows
how two coherent
point sources interfere
in an arbitrary point P.
They can interfere
constructively or
destructively,
depending on the
difference in traveled

13
5
 Chapter 6: Ultrasound imaging

Figure6.3 (a) The phon scale compensates for the effect of frequency on the perceived loudness in humans. By definition, x phon is equal
to x dB at a frequency of 1 kHz. 0 phon corresponds to the threshold of audibility for humans, and 120 phon to the pain threshold. An
increase of 10 phon corresponds approximately to twice the perceived loudness. (b) A-weighting. This curve is an approximation of the
isophones in the range 20–40 phons, shown in (a). It expresses the shift of the sound level (in dB) as a function of frequency to obtain an
approximation of

the perceived loudness

level in dB(A).

O1

O2

Figure6.4 Two coherent

point sources, originating
from positions O1 and O2
respectively, travel in all
directions, but only the
direction towards position
P is shown. When the
waves meet in P, they can
amplify each other, i.e.,
interfere constructively, or
depress each other, i.e.,
interfere destructively.
Maximal constructive
interference is the case
when δ = nλ with λ the
wavelength, while
1
complete )λ. destructive
interference happens when
δ = (n + 2

contributions of all
wavelets from all
point sources will
interfere
constructively because
their phase difference
13
6
 Chapter 6: Ultrasound imaging

is negligible. pressure close to

Moreover, at such a the source. Points
large distance, moving that are at least at a
away from the distance where all
symmetry axis does contributions start
not significantly to interfere
influence the phase constructively are
differences, and located in the so-
variations in maximal called farfield,
pressure occur slowly. whereas closer
However, when the points are in the
point of observation near field.
comes closer to the
source, contributions Attenuation
of different point Attenuation refers
sources interfere in a to the loss of
complex way because acoustic energy of
phase differences the ultrasonic
become significant. wave during
This results in fast propagation. In
oscillations of the tissues, attenuation
maximal is mainly due to
the conversion of
acoustic energy
into heat because
0 of viscosity. It
Sound level (dB)

–10 results in an
–20 exponential decay
–30 of the amplitude of
–40 the propagating
10
wave. 96
5 72
0 Typically,
48 Depth (mm) the
24
Lateral distance (mm) –5 attenuation is a
120
function of the
0 wave frequency.
Therefore, it is
Figure6.5 Simulated spatial
distribution of the maximal often modeled as a
pressure generated by a function of the
circular, planar source with
a diameter of 10 mm, form
transmitting a pulse with a
center frequency of 5 MHz. H(f , z) = e−αz ≡ e−α0f nz,
(Courtesy of Professor J.
D’hooge, Department of (6.9)
Cardiology. Reprinted with
permission of Leuven
University Press.) where f is the
frequency and z
the distance
13
7
 Chapter 6: Ultrasound imaging

propagated frequency (hence,

through the n = 1). The
medium with constant α0 can
attenuation thus be expressed
coefficient α, in Np/(cm MHz)
which is expressed or dB/(cm MHz).
in nepers (Np) per If α0 is 0.5 dB/(cm
centimeter. A MHz) a 2 MHz
neper is a wave will
dimensionless unit approximately
used to express a halve its amplitude
ratio. The value of after 6 cm of
a ratio in nepers is propagation. Some
given by ln(pz/p0) typical values of
where pz and p0 are α0 for different
the amplitudes of substances are
the wave at the given in Table 6.2.
distances z and 0,
respectively. Nonlinearity
In the The derivation of
literature, the unit the wave equation
dB/cm is often Eq. (6.2), assumes
used. According to that the acoustic
Eq. (6.8) the value pressure p is only
in decibels is an infinitesimal
given by 20 disturbance of the
log10(pz/p0). To use static pressure. In
expression (6.9) that case, the
the conversion linear wave
from dB/cm to equation can be
Np/cm needs to be derived, which
done by dividing α shows that any
by a factor 20 waveform
log10(e) = 8.6859. propagates
through a medium
It has been without changing
observed that, its shape (cf.
within the Figure 6.2).
frequency range However, if the
used for medical acoustic pressure
ultrasound increases, this
imaging, most approximation is
tissues have an no longer valid,
attenuation and wave
coefficient that is propagation is
linearly associated with
proportional to the distortion of the
13
8
 Chapter 6: Ultrasound imaging

waveform. This S
effect is visible in o
the frequency m
e
domain by the
generation of
t
higherharmonics(i.
y
e., p
integermultiplesoft i
heoriginal c
frequency). a
To illustrate l
this phenomenon,
a propagating v
wave was a
l
measured as a
u
function of time. e
Figure 6.6 s
represents these
recorded time o
signals and their f
spectra. The origin
of the time axis α
(time = 0) is the 0

moment at which
the pulse was f
generated. The o
r
distortion of the
waveform and the
d
introduction of i
harmonic f
frequencies f
increase with e
increasing r
pressure e
amplitude, which n
can be noticed by t
comparing the top
s
row with
u
T
b
a
b s
l t
e a
n
6 c
. e
2 s
Substance α0 (dB/(cm MHz))
13
9
 Chapter 6: Ultrasound imaging

Lung Physical meaning of A

Bone and B
Kidney The acoustic pressure
Liver p can be expressed as
Brain a function of the
Fat density ρ using a
Blood Taylor expansion
around the static
Water
density ρ0
the central row in the
diagram. Furthermore, p
because nonlinear
wave distortion is
induced during
propagation, the effect where is
increases with the acoustic density.
propagation distance. For small amplitudes
This is visible when of ρ (and,
comparing the bottom consequently, of p), a
and central rows in firstorder
the diagram, which approximation of the
show the previous equation can
measurement of an be used:
identical pulse at a
different distance
from the source. p .
The rate of (6.11)
harmonics generation
In this case, which is
at constant pressure
used to derive the
amplitude is different
linear wave equation,
for different media.
a linear relationship
The nonlinearity of a
exists between the
medium is described
by its nonlinearity
parameter B/A. Table
6.3 gives an overview
of some B/A values
for different biologic
tissues. The larger the
value of B/A is, the
more pronounced the
nonlinearity of the
medium.

14
0
 Chapter 6: Ultrasound imaging

Figure6.6 (a) The nonlinear propagation of a wave yields a distortion of the original waveform. (b) This distortion is visible in the
frequency domain as higher harmonics. The amount of distortion depends on both the amplitude (central versus top row) and the
propagation distance (bottom versus central row). (Courtesy of Professor J. D’hooge, Department of Cardiology. Reprinted with
permission of Leuven University

Press.) and

Table 6.3
Nonlinearity
B .
parameter (6.13)
B/A of some Eq. (6.10) can be
biologic rewritten as
tissues
Medium B/A

Pig blood 6.2

Spleen 7.8
Muscle 6.5
H2O 5.0
p
acoustic pressure and
the acoustic density.
However, if the .
amplitude of ρ is not (6.14)
small, the second- and
higher order terms The larger B/A is, the
cannot be neglected.
stronger the
Defining
nonlinearity effect.

A
(6.12)

14
1
 Chapter 6: Ultrasound imaging

Wave
propagatio
n in
inhomogen
eous media
Tissues are
inhomogeneous
media. When
inhomogeneities
are present,
additional
phenomena occur,
which can be
explained using
Huygens’ Figure6.7 Schematic
principle, as shown representation of
in Figure 6.7. This Huygens’ principle. The
concentric lines in this
states that any figure represent the
point on a wavefronts, i.e.,
wavefrontcanbeco surfaces where the
waves have the same
nsideredasthesourc phase. Any point on a
eofsecondary wavefront can be
waves, and the considered as the
source of secondary
surface tangent to waves and the surface
these secondary tangent to these
waves determines secondary waves
determines the future
the future position position of the
of the wavefront. wavefront.

A wavefront is a
surface where the
waves have the
same phase.

Reflection and
refraction
When a wave
propagating in a
medium with
density ρ1 and
sound velocity c1
meets another
14
2
 Chapter 6: Ultrasound imaging

medium with same direction as

density ρ2 and the incident wave.
sound velocity c2, Therefore, the
as illustrated in transmitted wave
Figures 6.8 and is also called the
6.9, part of the refracted wave.
energy of the wave From Eq. (6.15) it
is reflected and I1 Ir c1 = l 1.f
part is transmitted. c2 = l 2.f
The frequency of
reflected ui ur
boththereflectedan wavefront incident
drefractedwavesist wavefront
hesameasthe
c1
incident
frequency. Using 1 2 3 4 c2
Huygens’
principle, it can Figure6.8 Schematic
easily be shown representation of reflection
of an incident wave at a
that the angles of planar interface of two
the traveling different media. The
(planar) waves relationships between the
angles θi is given in Eq.
with respect to the (6.15).
planar interface
have the following
relationship,
which is Snell’s
law:

sin θi
sin θ

, (6.15)
c

with θi, θr and θt

the angles of
incidence,
reflection, and
transmission,
respectively. The
transmitted wave
does not
necessarily
propagate in the
14
3
 Chapter 6: Ultrasound imaging

I1 shown that these

c1 = l 1. f
amplitudes relatec as
=l .f 2 2

ui
Aincident
t 2Z2 cos
θi wavefront
T≡ =c +
1
1 2 3 4 (6.17)
c2
refracted
wavefront Ai Z2 cos
θi Z1 cos θt
u
andt

It Ar Z2 cos
−
Figure6.9 Schematic θi Z1 cos θt
representation of refraction R≡ = +
of an incident wave at a
planar interface of two , (6.18)
different media. The
relationships between the Ai Z2 cos
angles θi is given in Eq. θi Z1 cos θt
(6.15).

where Ai, Ar, and At are

the incident, reflected,
follows that and transmitted
amplitudes and Z1 and
Z2 the specific acoustic
impedances of the two
media. The
parameters T and R
. (6.16)
are called the
transmission
If c2 > c1 and θi > coefficient and the
sin−1(c1/c2), cos θt reflection coefficient
becomes a complex and relate as
number. In this case,
it can be shown that R = T − 1.
the incident and (6.19)
reflected waves are
out of phase.
Not only does the The reflection is large
direction of if Z1 and Z2 differ
propagation at the strongly, such as for
interface between two tissue/bone and
media change, but air/tissue transitions
also the amplitude of (see Table 6.1). This
the waves. In the case is the reason why in
of a smooth planar diagnostic imaging a
interface, it can be gel is used that
14
4
 Chapter 6: Ultrasound imaging

couples the transducer owing to local

and the patient’s skin. deviations of
Note that when a density and
wave propagates from compressibility.
medium 1 into Scatter reflections
medium 2, T and R therefore
are different than contribute to the
when the same wave signal, as is
travels from medium illustrated in
2 into medium 1. Both Figure 6.10.
coefficients can The smallest
therefore be assigned possible
indices to indicate in inhomogeneity is a
which direction the point and is called
wave propagates (e.g., a point scatterer. A
T12 or R21). point scatterer
The type of retransmits the
reflections discussed incident wave
in this paragraph are equally in all
pure specular directions as if it
reflections. They were a source of
occur at perfectly ultrasonic waves
smooth surfaces, (Huygens’
which is rarely the principle). Waves
case in practice. scattered in the
Reflections are opposite direction
significant in a cone to the incident
centered around the pulse are called
theoretical direction backscatter. The
θr. This is the reason characteristics of a
why a single finite scatterer can
transducer can be used be understood by
for the transmission considering it as a
and measurement of collection of point
the ultrasonic waves. scatterers. Since
Scattering each point scatterer
From the previous retransmits the
discussion, it must received pulse in
not be concluded all directions, the
that reflections scattered pulse
only occur at from a finite
tissue boundaries scatterer is the
(e.g., blood– interference
muscle). In between the
practice, individual wavelets from the
tissues are constituting point
inhomogeneous scatterers.
14
5
 Chapter 6: Ultrasound imaging

Obviously, this wavelength,

interference allcontributionsinte
pattern depends on rfereconstructively
the shape and size , independently of
of the scatterer. the shape of the
Becausethispattern scatterer or of the
istheresultoftheinte point of
rferenceofa very observation P (as
large number of long as it is far
coherent enough from the
(secondary) scatterer). This is
sources, it is also illustrated
known as the schematically in
diffraction pattern Figure 6.11(a). On
of the scatterer. the other hand, if
If the scatterer the size of the
is much smaller objectiscomparable
than the tothewavelength,
thereisaphase
4 20

3 15

2
10
Amplitude (mV)

Amplitude (mV)

1
5
0
0
–1
–5
–2
–10
–3

–4 –15

–5 –20
20 30 40 50 60 70 80 90 100110 20 30 40 50 60 70 80 90 100110
Time (ms) Time (ms)
(a) (b )

Figure6.10 (a) Reflected signal as a function of time for a

homogeneous object in water. Obviously, a large
reflection occurs at the interfaces between the two
matters. The other apparent reflections within the water
and object are caused by acquisition noise (note their
small amplitude). (b) Reflected signal as a function of time
for an inhomogeneous object in water. The reflections
show an exponential decay.
Deeper regions in the scattering object have a smaller
amplitude. This is due to attenuation. Note the different
scales in both diagrams.
(Courtesy of Professor J. D’hooge, Department of Cardiology.
Reprinted with permission of Leuven University Press.)

14
6
 Chapter 6: Ultrasound imaging

t
P P
t
t

t
t

(a) (b)

Figure6.11 A scatterer can be represented by a collection

of point scatterers. Each point scatterer retransmits the
incident pressure field in all directions. (a) These
wavelets interfere constructively at the point P of
observation if the scatterer is much smaller than the
wavelength. (b) They interfere in a complex manner
when the size of the scatterer is comparable to or larger
than the wavelength.
shift between the
retransmitted
Wav
wavelets, and the e
interference
pattern depends on pro
the shape of the
scatterer and the
pag
point of atio
observation (see
Figure 6.11(b)). n
and
moti
on:
the
Dop
pler
effe
ct
If an acoustic
source moves
relative to an
14
7
 Chapter 6: Ultrasound imaging

observer, the =−
fD = fR − fT +
frequencies of the
observed and fT.
transmitted waves
are different. This
(6.20
is the Doppler
effect. A well- ) c
known
va
exampleisthatofaw
histlingtrainpassin
Proof of Eq. (6.20)
ganobserver. The
The position Ps(t)
observed pitch of
of the start point of
the whistle is
the transmitted
higher when the
pulse at time t can
train approaches
be written as
than when it
moves in the other
direction. Ps(t) = ct,
Consider the (6.21)
l
schematic
representation of a
v
transducer, a
u
transmitted pulse,
Transducer
and a point Pe Ps P va
scatterer in Figure
6.12. Assume that
the scatterer d 0 + va . t
moves away from
Figure6.12 Geometry used
the static to derive an expression for
transducer with an the frequency shift due to
axial velocity the motion of the scatterer
(i.e., the Doppler
component va = |v| frequency).
· cos θ. If fT is the
frequency of the
where c is the sound
pulse transmitted
velocity within the
by the transducer,
medium and t the time
the moving
since the pulse was
scatterer reflects
transmitted. The
this pulse at a
position P(t) of the
different frequency
scatterer can be
fR. The frequency
written as
shift fD = fR − fT is
the Doppler
P(t) = d0 + vat,
frequency and can
(6.22)
be written as

2va
14
8
 Chapter 6: Ultrasound imaging

where d0 is the distance =

from the transducer to
the scatterer at time t = c
0. t
The start point of the
−
ultrasonic pulse meets
the scatterer when Ps(t) λ
(6.24)
= P(t), say at time tis.
=
Hence, Ps(tis) = P(tis). c(t
From Eqs. (6.21) and −
T).
(6.22) it follows that

ctis = d0 + Point Pe meets the

vatis
scatterer if Pe(t) = P(t),
d
0 say at time tie. Hence,

(6. Pe(tie) = P(tie). From

23
) Eqs. (6.24) and (6.22)
tis
= it then follows that

c(tie − T) = d0 + vatie
.c d
−
cT
va
tie = 0
+ c − va
Without loss of c
generality, assume that
the pulse length equals = tis + −
one wavelength λ. If T
T
is the period of the
transmitted pulse, the .
position Pe(t) of the
(
end point can be
written as 6
.
P
e 2
( 5
t
) )
c
14
9
 Chapter 6: Ultrasound imaging

+ trs and te = tie + tre

v respectively. Hence,

a
ts =
2tis
The scatterer reflects
(6.27)
the pulse back to the
receiver. The position te = 2tie − T.
where the points Ps
and Pe meet the Consequently,
scatterer are
thedurationTR = te−ts
respectively Ps(tis) and
Pe(tie). These are also ofthereceived pulse
the distances these can easily be obtained
points have to travel
back to the transducer. from Eq. (6.25):
The corresponding c
travel times trs and tre TR = 2
are

Ps(tis)
−
t
r
s
T

=
−
=

t T
i
s

c c
Pe(tie
) v
tre = = tie − T, a

(6.26) c

which have to be
(
added to tis and tie
6.2
respectively to 8)
calculate the travel T.
time back and forth of
Ps and Pe, that is, ts = tis Substituting TR by 1/fR
and T by 1/fT in this
15
0
 Chapter 6: Ultrasound imaging

equation, yields the

Doppler frequency fD
Generation
and
2va
fD = fR − fT
=−
+ fT.
detection
of
(6.29) c ultrasound
va Ultrasonic waves
are both generated
In practice, the and detected by a
velocity of the piezoelectric
scatterer is much crystal, which
smaller than the deforms under the
velocity of sound influence of an
and the Doppler electric field and,
frequency can be vice versa, induces
approximated as an electric field
over the crystal
2|v| cos θ after deformation.
fD ≈ − This crystal is
fT. (6.30) embedded in a so-
c called transducer
that serves both as
For example, if a a transmitter and
scatterer moves as a detector. Two
away from the piezoelectric
transducer with a materials that are
velocity of 0.5 m/s often used are PZT
and the pulse (lead zirconate
frequency is 2.5 titanate) and PVDF
MHz, the Doppler (polyvinylidene
shift is fluoride), which
approximately are both polymers.
−1.6 kHz. Note If a
that for θ = 90◦ the piezoelectric
crystal is driven
Doppler frequency
with a sinusoidal
is zero.
electrical signal, its
surfaces move, and
a compression
wave at the same
frequency is
generated and
propagates through
the surrounding
media. However,
15
1
 Chapter 6: Ultrasound imaging

Eqs. (6.17) and image side.

(6.18) show that Therefore,
the transmission appropriate
and reflection materials with
coefficients at the totally different
interface between acoustic
two media depend impedance than
on the acoustic that of the crystal
impedances of the are used as a
media. Therefore, backing at the
part of the energy other surface
produced by the (Figure 6.13). As
transducer is such, almost all
reflected inside the energy is reflected
crystal and into the crystal. At
propagates toward the front side of
the opposite the crystal, as
surface. Because muchenergyasposs
this reflected ibleshouldbetrans
(compression) mittedintothe
wave in turn Figure6.13
induces electrical Schematic
illustration of the
fields that interfere cross-section of an
Matching
with the driving
Backing

ultrasonic
Crystal

transducer, which
electrical force, it consists of four
can be shown that important elements
the amplitude of (i.e., a backing layer,
electrodes, a
the vibration is piezoelectric crystal,
maximal when the and a matching
thickness of the layer).
crystal is exactly
half the
Electrodes
wavelength of the
induced wave. This
phenomenon is
called resonance, medium. However,
and the because the
correspond- acoustic
ing frequency is called impedance of
the fundamental solids is very
resonance frequency. different from that
Obviously, as of fluids (which is
much of the a good model for
acoustic energy as biologic tissue),
possible should part of the energy
emerge from the is reflected into
crystal through the the crystal again.
surface on the This problem is
15
2
 Chapter 6: Ultrasound imaging

solved by using a
so-called matching
Gray
layer (Figure scale
6.13),√
which has an imaging
acoustic
impedance equal
Data
to ZcZt, with Zc and acquisitio
Zt being the
acoustic n
impedance of the Insteadofapplyinga
crystal and the continuouselectric
tissue, alsignaltothe
respectively. It can crystal, pulses are
be shown that if used to obtain
this layer has a spatial
thickness equal to information.
an odd number of Data acquisition is
quarter done in three
wavelengths, different ways.
complete
A-mode
transmission of
energy from the Immediately after
the transmission of
crystal to the
the pulse, the
tissue can be
transducer is used
obtained.
as a receiver. The
An ultrasonic
reflected (both
transducer can
specular and
only generate and scattered) waves
receive a limited are recorded as a
band of function of time.
frequencies. This An example has
band is called the already been
bandwidth of the shown in Figure
transducer. 6.10. Note that
time and depth are
equivalent in
echography
because the sound
velocity is
approximately
constant
throughout the
tissue. In other
words, c
multiplied by the
travel time of the
15
3
 Chapter 6: Ultrasound imaging

pulse equals twice B-mode

the distance from A 2D image (e.g.,
the transducer to Figure 6.16) can be
the reflection obtained by
point. This translating the
simplest form of transducer between
ultrasound two A-mode
imaging, based on
the pulse–echo
Line number

Amplitude (volt)
principle, is called
A-mode
(amplitude)
imaging. The
detected signal is 0.02

often called the 0 25

radiofrequency 0.02 20
(RF) signal 0
15
because the 40 Line number
60 10
frequencies 80 Depth (mm)
5
involved are in the 100
120 0
MHz range and
correspond to the Figure6.14 Repeated A-
frequencies of mode measurement yields
M-mode imaging. (Courtesy
radio waves in the of Professor J. D’hooge,
electromagnetic Department of Cardiology.
spectrum. Reprinted with permission
of Leuven University Press.)
M-mode
The A-mode Line number
measurement can be
repeated. For a static
transducer and object,
all the acquired lines
are identical, but if the
object moves, the
signal changes. This
kind of imaging is
called M-mode
(motion) imaging and Depth
yields a 2D image Figure6.15 M-mode image
with depth and line of the heart wall for
number as the assessment of cardiac wall
motion during contraction.
dimensions (see The black region is blood,
Figures 6.14 and the bright reflection is the
pericardium (i.e., a
6.15). membrane around the
heart), and the gray region
in between is the heart
muscle itself.

15
4
 Chapter 6: Ultrasound imaging

(Courtesy of Professor J. the waves can

D’hooge, Department of
Cardiology. approach the heart

Transducer
Tra
nsd
uce
r
Object
Object cer
du
ns
Tra

Transducer

(a) (b)

Figure6.16 (a) B-mode image of a fetus. The dark region is the uterus, which is filled with fluid. (Courtesy of Professor M. H. Smet,
Department of Radiology) (b) B-mode image of a normal heart in a four-chamber view showing the two ventricles (LV left ventricle; RV
right ventricle), the two atria (LA left atrium; RA right atrium) and the origin of the aorta (outflow tract). Besides the anatomy of the
whole heart, the morphology of the valves (e.g., mitral valve) can be visualized. (Courtesy of the Department of Cardiology.)

Figure6.17 B-mode image acquisition can be done by

either translating (a) or tilting (b) the transducer.
(Courtesy of Professor J. D’hooge, Department of
Cardiology. Reprinted with permission of
Leuven University Press.)
Reprinted with permission only through the small
of Leuven University Press.) space between the
acquisitions. This is ribs. This space is
illustrated in Figure often called the
6.17(a). This kind of acoustic window.
imaging is called B- Because the acoustic
mode imaging, where window for the heart
B stands for brightness is relatively small, the
(see also p. 140). If translation technique
this measurement is
described above
repeated over time, an
cannot be applied to
image sequence is
cardiac applications.
obtained.
A possible solution is
Because bone has
to scan a sector by
a high attenuation
tilting the transducer
coefficient,
rather than translating
transmission of sound
it (see Figure 6.17(b)).
through bone is
The same
minimal. For example,
imaging modes are
15
5
 Chapter 6: Ultrasound imaging

also used for Allofthesestepsarenow

second harmonic brieflydiscussed.
imaging. The
difference with Filtering
traditional imaging First, the received
is that the complete RF signals are
bandwidth of the filtered in order to
transducer is not remove high-
used during frequency noise.
transmission but In second
only a low- harmonic imaging,
frequency part. the transmitted
Higher harmonics low-frequency
are generated band is also
during wave removed, leaving
propagation and only the received
are detected with high frequencies
the remaining in the upper part
high-frequency of the bandwidth
part of the of the transducer.
sensitive The origin of these
bandwidth of the frequencies, which
transducer. The were not
bandwidth used transmitted, is
during nonlinear wave
transmission can propagation (see p.
be changed by 132). Figure 6.18
modifying the shows
properties of the fundamental and
electrical pulses second harmonic
that excite the images of the
heart.
Image
crystals.

reconstruc Envelope
detection
tion Because the very
Reconstructing fast fluctuations of
ultrasound images the RF signal (as
based on the acquired illustrated in
RF data as shown in Figure 6.19(a)) are
Figure 6.14, involves not relevant for
the following steps: gray scale
filtering, envelope imaging, the high-
detection, attenuation frequency
correction, log- information is
compression, and scan removed by
conversion. envelope
15
6
 Chapter 6: Ultrasound imaging

detection. Usually To construct an M-

this is done by mode image, the same
means of a procedure with a static
quadrature filter or transducer is applied.
a Hilbert The result is a gray
transformation. A value image with
detailed depth and time as the
discussion, dimensions, as
however, is illustrated in Figure
beyond the scope 6.15.
of this text.
Figure 6.19(a)
shows an example of Attenuation
an RF signal and its correction
envelope. If each Identical structures
amplitude along the should have the same
envelope is gray value and,
represented as a gray consequently, the
value or brightness, same reflection
and different lines are amplitudes. However,
scanned by translating the amplitude of the
the transducer, a B- incident and reflected
mode (B stands for wave decreases with
brightness) image is depth because of
obtained. Figure attenuation of the
6.19(b) shows an acoustic energy of the
example. Bright pixels ultrasonic wave
correspond to strong during propagation
reflections, and the (see Eq. (6.9)). To
white lines in the compensate for this
image represent the effect, the attenuation
two boundaries of the is estimated. Because
scanned object. time

RV LV

RA
LA

Figure6.18 B-mode image of the heart in fundamental (a) and

second harmonic (b) imaging modes. The two ventricular and
the two atrial cavities are shown as dark regions. The heart
muscle itself is situated between the cavities and the bright
specular reflection. Second harmonic imaging yields a better
15
7
 Chapter 6: Ultrasound imaging

image quality in corpulent patients. (Courtesy of the

Department of Cardiology.)

(a) (b)

40
5
30 10
Lateral displacement (mm)
Amplitude (mV)

20 15

10 20
25
0
30
–10
35
–20 40
–30 45

–40 50
0 2 4 6 8 10 2 4 6 8 10 12 14 16 18 20
Time (ms) Time ( s)

(a) (b)

Figure6.19 Plotting the amplitudes of the envelope (a) as gray

values yields an ultrasound image (b). (Courtesy of the
Department of Cardiology.)
Figure6.20 Processing the original gray
250 5 values as indicated in
(a) results in a better
10 visualization of the scatter reflections
(b).
Lateral displacement (mm)

200
Processed gray value

15 This is an example of a gray level

20 transformation, as discussed in Chapter 1,
150 p. 4. (Courtesy of the Department of
25
Cardiology.)
30
100
35
40
50
45
50
0 2 4 6 8 10 12 14 16 18 20
0 50 100 150 200 250
Original gray value Time ( s)

(a) (b)
and depth are linearly ultrasound scanners
related in echography, therefore enable the
attenuation correction user to modify the
is often called time gain manually at
gain compensation. different depths.
Typically, a simple
model is used – for Log-compression
example, an Figure 6.19(b) mainly
exponential decay – shows the specular
but in practice several reflections. However,
tissueswithdifferentatt the scatter reflections
enuationpropertiesarei are almost invisible.
nvolved. Most Thereasonisthelargedi
15
8
 Chapter 6: Ultrasound imaging

fferenceinamplitudebe
tween the specular
Acquisition
and the scatter and
reflections, yielding a
large dynamic range. reconstructio
In order to overcome
this problem, a
n time
suitable gray level To have an idea of the
transformation can be acquisition time of a
applied (see Chapter typical ultrasound
1, p. 4). Typically, a image, a simple
logarithmic function is calculation can be
used (Figure 6.20(a)). made. Typically, each
The log-compressed line in the image
version of the image corresponds to a depth
in Figure 6.19(b) is of 20 cm. Because the
shown in Figure velocity of sound is
6.20(b). The scatter or approximately 1540
speckle can now m/s and the travel
easily be perceived. distance to and from
Note that different the transducer is 40
tissues generate cm, the acquisition of
different speckle each line takes 267 µs.
patterns. A typical image with
120 image lines then
Scan conversion requires an acquisition
If the image is time of about 32 ms.
acquired by tilting the The reconstruction of
transducer instead of the images can be
translating it, samples done in real time.
on a polar grid are Consequently, a
obtained. Converting temporal resolution of
the polar into a 30 Hz (i.e., 30 images
rectangular grid needs per second) can be
interpolation. This obtained. This
process is called scan temporal
conversion or sector resolutioncanbeincrea
reconstruction. sedatthecostofspatialre
solution by decreasing
the number of scan
lines. However,
current clinical
scanners are able to
acquire multiple scan
lines simultaneously
with little influence on
the spatial resolution
15
9
 Chapter 6: Ultrasound imaging

(see p. 150 below). Doppler is the

This way, frame rates only exception
of 70–80 Hz can be to the pulse–
obtained. echo principle
for ultrasound
data
Doppler acquisition. It
does not yield
imaging spatial (i.e.,
Doppler imaging is depth)
a general term information.
used to visualize • Pulsedwave(P
velocities of W)Doppler
moving tissues. Pulsed waves
Data acquisition are transmitted
and reconstruction along a
are different from particular line
gray scale through the
imaging.
tissue at a
Furthermore,
constant pulse
theDopplerprincipl
repetition
eisnotalwaysused.
frequency
Data (PRF).
However,
acquisition rather than
In Doppler imaging acquiring the
data acquisition is complete RF
done in three different signal as a
ways. function of
time, as in the
• Continuous
M-mode
wave (CW)
acquisition (see
Doppler A
p. 138), only
continuous
one sample of
sinusoidal
each reflected
wave is
pulse is taken
transmitted by
at a fixed time,
a piezoelectric
the so-called
crystal, and the
range gate,
reflected signal
after the
is received by a
transmission of
second crystal.
the pulse.
Usually, both
Consequently,
crystals are
information is
embedded in
obtained from
the same
one specific
transducer. CW
spatial position.
16
0
 Chapter 6: Ultrasound imaging

• Color flow frequency of the

(CF) imaging received wave fR is
This is the compared with
Doppler that of the
equivalent of transmitted wave
the B-mode fT. Equation (6.30)
acquisition (see gives the relation
p. 138). between the
However, for Doppler frequency
each image fD = fR − fT and the
line, several velocity va of the
pulses tissue. Note that
(typically3– only the axial
7)insteadofonea component of the
retransmitted. object’s motion
Theresult is a can be measured
2D image in this way.
which the Cardiac and
velocity blood velocities
information is cause a Doppler
visualized by shift in the sonic
means of color range. Therefore,
superimposed the Doppler
onto the frequency is often
anatomical made audible to
gray scale the user. A high
image. pitch corresponds
to a high velocity,
whereas a low
Reconstruc pitch corresponds
to a low velocity.
tion If the received
From the acquired signal is
data the velocity subdivided into
must be calculated segments, the
and visualized. frequency
This is different spectrum at
for each of the subsequent time
acquisition modes. intervals can be
obtained from their
Continuous wave
Fourier transform.
Doppler
The spectral
To calculate the
amplitude can then
velocity of a
be encoded as a
scattering object in
gray value in an
front of the
image. This picture
transducer, the
is called the
16
1
 Chapter 6: Ultrasound imaging

spectrogram or other words, a

sonogram. compromise has to
Typically, be made between
scatterers with the
different velocities
are present in one
segment, yielding
a range of Doppler
shifts and a broad
instead of a sparse
peaked spectrum.
Moreover,
segmentation of
the received signal
corresponds to a
multiplication with
a rectangular
function and a Figure6.21 CW Doppler
spectrogram showing the
convolution of the velocity profile of the blood
true spectrum with flow through a heart valve.
a sinc function. (Courtesy of the
Department of Cardiology.)
Consequently, the
spectrum appears
smoother than it
actually should be. velocity resolution and
This kind of temporal resolution of
broadening is the spectrogram.
called intrinsic
broadening Pulsed wave doppler
because it has no Pulsed wave (PW)
physical origin but Doppler does not
is due purely to make use of the
signal processing. Doppler principle.
An example of a Instead, the received
continuous wave signal is assumed to
spectrogram is be a scaled, delayed
given in Figure replica with the same
6.21. It is clear that frequency (i.e., fR = fT)
a compromise has as the transmitted
to be made pulse. Effects such as
between the diffraction and
amount of intrinsic nonlinearity are also
spectral neglected. Assume a
broadening and the transmitted sinusoidal
time between two wave
spectra in the
spectrogram. In
16
2
 Chapter 6: Ultrasound imaging

p(t) = ymax
Line number
(6.31)

Amplitude (volt)
The received signal
then is
0.02

0 25

0.02 20
20
. (6.32) 15
40 Line number
range 60 10
The interval t is the 80 Depth (mm)
gate tR 5
time between the 100
transmission and the 120 0

reception of the pulse Figure6.22 Pulsed wave

and depends on the Doppler uses the M-mode
acquisition scheme (see
distance d from the Figure (6.14)) and samples
transducer to the the subsequent reflected
pulses at a fixed range gate
scatterer, i.e., tR to calculate the Doppler
c. The PW frequency fD.
Doppler system takes
only one sample of
each of the received
pulses at a fixed range
gate tR (Figure 6.22),
that is,

If the scattering
object moves away
from the transducer
with a constant axial
velocity v a, the
distance d increases
between subsequent
pulses with vaTPRF,
where TPRF is the pulse
repetition period, TPRF
=

16
3
 Chapter 6: Ultrasound imaging

sinusoidal function
US pulse moving scatterer
with frequency

2va
fD = −

samples at tR
fT, (6.35)
c
which is exactly
the Doppler
frequency defined
in Eq. (6.30).
tR Hence, the velocity
reflected pulsesof the scattering
object at a specific
Figure6.23 Schematic
representation of the PW depth, defined by
Doppler principle. Pulses the range gate, can
are transmitted at a fixed be calculated from
pulse repetition frequency
(PRF). They are reflected at the sampled signal
a scatterer in motion. sj(tR). As in CW
Because of this motion, the
reflected pulses are Doppler, this
dephased. Measuring each signal can be made
reflected pulse at the range
gate tR yields a sampled
audible or can be
sinusoidal signal with represented as a
frequency fD. spectrogram.
Often, the
spectrogram is
1/PRF (see Figure displayed together
6.23). Consequently, with a B-mode
Eq. (6.33) can be gray scale image in
written as which the position
sj of the range gate is
indicated. Such a
combined scan is
called a duplex.

Note that a PW
Doppler system as
TPRF described above is
(6.34) not able to detect
the direction of
where sj(tR) is the motion of the
sample of the jth scattering object.
pulse. Hence, the To obtain
values sj(tR) in Eq. directional
(6.34) are samples of information, not
a slowly time-varying one but two
samples, a quarter
16
4
 Chapter 6: Ultrasound imaging

of a wavelength or signal with

less apart, have to frequency fD (see
be taken each time Eq. (6.34)) color
because this flow imaging
completely calculates the
determines the phase shift
motion direction of between two
the waveform as subsequent
illustrated received pulses.
schematically in Equation (6.34)
Figure 6.24. shows that this
Similarly to phase shift can
CW Doppler, the then be used to
received (sampled) calculate the
signal is velocity va:
subdivided into
segments and their
frequency
spectrum is 2vaTPR
F
obtained by
fT
calculating the
Fourier transform.
.
The result is also
visualized as a
(6.
spectrogram.
36)
An example is shown
c
in Figure 6.25.
The phase shift φ
Color flow imaging
can be derived by
Similar to pulsed
sampling two
wave (PW),
subsequent pulses
Doppler color flow
at two specific
(CF) imaging uses
time instances tR1
ultrasonic pulses
and tR2. It can
and makes the
easily be shown
same assumption
that two samples
that the received
of a sinusoid
signal s(t) is a completely
scaled, delayed determine its phase
replica with the given that they are
same frequency no more than a
(i.e., fR = fT) as the quarter of a
transmitted wavelength apart
(sinusoidal) pulse. (see Figure 6.24).
However, instead Consequently, two
of calculating va such unique
from samples of a couples of samples
16
5
 Chapter 6: Ultrasound imaging

are sufficient to that it does not suffer

determine from aliasing (see
completely the Appendix A, p. 230).
phase shift φ Moreover,
between two crosscorrelation is
subsequent pulses. not limited to one-
Anotherwaytocal dimensional signals.

? tR ! tR1 tR2

tR tR1 tR2

? ?
tR tR1 tR2

(a) (b)

Figure6.24 If a single sample is acquired at the range gate, no directional information is obtained (a). However, if a second sample is
acquired slightly after the first one, the direction of motion is uniquely determined since a unique couple of samples within the cycle is
obtained (b).

culateφ It can also be applied

isbycross-correlation to the subsequent
of the signals ultrasound images of
received from both a time sequence.
pulses. This method This has the
has the advantage advantage that the
16
6
 Chapter 6: Ultrasound imaging

real velocity |v| from the acquired

instead of the axial RF signals.
velocity va = |v| cos Typically, red
θ is calculated. represents
However, velocities toward
crosscorrelation is the transducer, and
not commonly blue the opposite.
available yet in The color flow
clinical practice. technique can also be
In practice, the applied to data
measurements are acquired along a
noisy. To increase single image line. In
the accuracy, more that case, theM-
than two pulses modegrayscaleimagec
(typically three to anbedisplayedtogether
seven) are used with the estimated
and the results are local velocities. This
averaged. By is color flow M-mode.
dividing the whole
acquired RF line
into range gates in
which a few
samples are taken,
this method
permits
calculations of the
local velocity
along the complete
line (i.e., for all
depths). This
process can be
repeated for
different lines in
order to obtain a
2D image.
Usually, the
velocity
information is
color coded and
displayed on top of
the conventional
gray scale image
(see Figure 6.26,
for example),
which can be
reconstructed
simultaneously
16
7
 Chapter 6: Ultrasound imaging

(a) (b)

Figure6.26 (a) Using color Doppler techniques, blood flow within the ventricles can be visualized. This image shows the flow in a normal
left ventricle at the beginning of diastole. Red colors represent flow toward the transducer, coming from the left atrium through the
mitral valve and into the left ventricle. Blue colors show the blood within the left ventricle flowing away from the transducer toward the
aorta. (b) Doppler techniques can be used to acquire the slower, regional velocities of the heart muscle itself. Local velocities in the
direction of the transducer are represented in red, and velocities away from the transducer are in blue. (Courtesy of the Department of
Cardiology.)

Acquisition
and
reconstructio
n time
Continuous wave and
pulsed wave Doppler
require a long
transmission of either
CW or PW ultrasound
in tissue.
Reconstruction then
consists of a simple
Fourier transform,
which can be done in
real time. In practice,
changes in velocities
over time are
investigated, and the
received signal is
subdivided into small
(overlapping)
segments whose
spectral amplitudes
constitute the columns
of a spectrogram.
Figure 6.21 shows an
example of a
16
8
 Chapter 6: Ultrasound imaging

spectrogram of the
heart. One such
Image
complete spectrogram quality
is typically acquired
in 3 to 4 seconds, Spatial
which corresponds to
3 to 4 heart cycles.
resolution
In color flow The spatial resolution
imaging, a few pulses in ultrasound imaging
are sent along each distinguishes the
image line. This axial, lateral, and
means that the time elevation resolution,
needed to acquire an i.e.,
theresolutioninthedire
image is the time
ctionofwavepropagati
required to obtain a
on, the
Bmode gray scale
y
image times the
number of pulses sent
along each line. If, for
example, three pulses
per line are used, an x

Elevation
image is obtained in
z
approximately 100
ms. Velocity
calculation can be
done in real time, Figure6.27 Schematic
representation of the axial,
resulting in a frame lateral and elevation
rate of 10 Hz. To directions.
increase this
framerate,
thefieldofview(FOV)i
resolution
susuallydecreased.
perpendicular to
The size of the FOV
the axial direction
can be selected
within the image
independent of the
plane, and the
FOV of the gray scale
resolution
image. This way, the
perpendicular to
velocity calculations
the image plane
are limited to the
(see Figure 6.27).
regions that contain
relevant velocity
information. Axial resolution
Resolution can be
expressed by the
PSF, which
determines the
minimum distance
16
9
 Chapter 6: Ultrasound imaging

between Hence, the resolution

neighboring details can be improved by
that can still be increasing the
distinguished as transmitted frequency
separate objects. In (see for example
the axial direction, Figure 6.41(a) below).
the width x of the However, because the
PSF depends on attenuation increases
the duration T of with higher
the transmitted frequencies (see Eq.
pulse. (6.9)), the penetration
Becausethepulseha depth decreases.
stotravelbackandfo Hence, a compromise
rth, an object point has to be made
does not between spatial
contaminate the resolution and
received signal penetration.
ofaneighboringpoi In Doppler
ntifthedistancex imaging, the
betweenthem is continuous wave
larger than c 2 (CW) mode does
(Figure 6.28), not yield spatial
which is a measure information. The
of the resolution. axial resolution of
A typical 2.5 MHz pulsed wave (PW)
transducer has an Figure6.28
Schematic
axial resolution x representation of a
of approximately 2∆x c.∆T pulse reflecting at
0.5 mm. two surfaces x
apart. The reflected
While in MRI a pulses can be
small bandwidth and, resolved if
c.∆T .
consequently, a long
RF pulse for slice
selection is needed to
obtain the best
resolution (see c.∆T ∆x
Chapter 4, p. 73), in and color flow
ultrasound a short (CF) imaging
pulse with large theoretically can
bandwidth is needed. be the same as for
For technical reasons gray scale
beyond the scope of imaging. However,
this text, this it can be shown
bandwidth is limited that the velocity
by the transducer and resolution is
is proportional to the improved by
central frequency. increasing the
17
0
 Chapter 6: Ultrasound imaging

pulse length. lateral resolution is

Consequently, PW best.
and CF Doppler Figure 6.29 shows
systems have to an example of the
make a sound field produced
compromise, and by a planar and a
in practice the curved transducer.
axial resolution is Figure 6.30(b)
somewhat illustrates the finer
sacrificed to the speckle pattern
velocity resolution. obtained with a
We note here that concave crystal as
no such compared with Figure
compromise has to 6.30(a) produced with
be made when a planar crystal. Note
cross-correlation is that focusing can also
used for be obtained
reconstruction (see electronically by
p. 143). However, means of a phased-
cross-correlation is array transducer, as
not available in explained on p. 150
clinical practice below.
today. It can further be
shown that the beam
Lateral and elevation width can be reduced
resolution and, consequently, the
lateral resolution
The width of the PSF
improved by
is determined by the
increasing the
width of the ultrasonic
bandwidth and the
beam, which in the
central frequency of
first place depends on
the transmitted pulse.
the size and the shape
Because gray scale
of the transducer.
images are acquired
Unlike a planar
with shorter pulses
crystal, a concave
than PW and CF
crystal focuses the
Doppler images (see
ultrasonic beam at a
previous paragraph)
certain distance. At
their lateral resolution
this position, the beam
is better. On the other
is smallest and the
hand, the lateral

17
1
 Chapter 6: Ultrasound imaging

0.04 0.04

0.02

Amplitude (V)
0.02
Amplitude (V)

0 0

–0.02 –0.02

–0.04 –0.04
51.8
7 50.8
50.8 3.5 3
0 0
–3.5 Depth (mm) –3
Depth (mm)
6
49.8 –7 Lateral distance (mm)
49.8 –6 Lateral distance (mm)

(a) (b)

Figure6.29 Experimentally measured lateral pressure field of

a planar (a) and concave (b) circular transducer with a radius
of 13 mm and a central frequency of 5 MHz. Note the smaller
shape in (b), which is due to focusing.
Lateral displacement (mm)

Lateral displacement (mm)

5 5
10 10
15 15
20 20
25 25
30 30
35 35
40 40
45 45
50 50
Figure6.30 Image obtained with (a) a planar and (b) a
concave transducer. Because of beam focusing, the lateral
resolution in (b) is clearly superior to that in (a). (Courtesy
of Professor J. D’hooge, Department of Cardiology.
Reprinted
with permission of Leuven University
Press.)

2 4 6 8 10 12 14 16 18 20 2 4 6 8 10 12 14 16 18 20
Time (s) Time (s)

(a) (b)

resolution of pulsed Typically, the

Doppler systems is lateral and elevation
better than that of a resolution in the focal
continuous wave region is on the order
(CW) system, which of a few millimeters
has a smaller (i.e., an order of
bandwidth. magnitude worse than
17
2
 Chapter 6: Ultrasound imaging

the resolution in the

axial direction).
Image
Outside this region the contrast
beam width increases Strongly reflecting
and the resolution structures, such as
becomes even worse. calcifications or tissue
interfaces, yield bright
reflections and are
Noise called echogenic
The noisy pattern as asopposedtohypogeni
observed in Figure c structureswithweak
6.30 is almost reflections, such as
completelyduetoscatte blood. The received
rreflections. signal is not only due
Acquisitionnoise can to specular reflections
be neglected but also to scatter. The
compared with this large-amplitude
so-called speckle difference between
noise. It can be shown the specular and the
that if sufficient scatter reflections
scatterers are present yields a large dynamic
per unit volume and if range. Typically, a
their position is logarithmic function
completely random, is used to overcome
the SNR equals 1.92. this problem (Figure
This is poor compared 6.20).
with other imaging It is important to
modalities. However, note that in ultrasound
the speckle pattern, imaging the degree of
here defined as noise, perceptibility of
enables the user to tissues is not
distinguish different
tissues from each
other and thus
contains useful
information.
Therefore, defining (a) (b)
the speckle pattern as Figure6.31 (a) A
noise is not very focusing transducer
relevant. produces side lobes. (b)
The point scatterer in
front of the transducer
appears several times
in the reconstructed
image, once for each
lobe.

17
3
 Chapter 6: Ultrasound imaging

only defined by the scatterer is

contrast (i.e., the illustrated in
difference in Figure 6.31.
brightness in Because of the side
adjacent regions of lobes, two
the image), but scatterers wrongly
also by the appear in the
difference in image.
speckle pattern or
texture. Reverberations
If a reflected wave
arrives at the
Gray scale transducer, part of
the energy is
image converted to
electrical energy
artifacts and part is
Side lobes reflected again by
A closer look at the transducer
the pressure field surface. This latter
produced by a part starts
focused transducer propagating
(see Figure through the tissue
6.29(b)) reveals in the same way as
that the lateral the original pulse.
pressure profile That means that it
shows a main lobe is reflected by the
and side lobes. The tissue and detected
amplitudes of again. These
these side lobes higher order
are much smaller reflections are
but can called
nevertheless reverberations and
introduce image give rise to
artifacts. The phantom patterns
reflections from if the amplitude of
the side lobes can the wave is large
contribute enough (see Figure
significantly to the 6.41 below).
reflected signal Because the length
and introduce of the completed
information from trajectory for a
another direction reverberation is a
in the received multiple n of the
signal. An extreme distance d between
situation of a transducer and
single point
17
4
 Chapter 6: Ultrasound imaging

tissue, they appear Proof of Eq. (6.37)

at a distance n · d.

Doppler
image
artifacts
Aliasing
A common artifact
of pulsed Doppler
methods (PW and
CF) is aliasing
(see Figure 6.32).
Aliasing is due to
undersampling
(see Appendix A,
p. 230). The
principle is shown
schematically in
Figure 6.33.
The following
constraint between the
range d0 and the
velocity va can be
deduced:

c
2
|va|.

(6
.3
7)
d

For example, given a

range gate at a depth
of 6 cm and an
ultrasonic frequency
of 5 MHz, the velocity
that can be measured
without aliasing is
restricted to
approximately 1 m/s.
17
5
 Chapter 6: Ultrasound imaging

According to the
Nyquist criterion, the
sampling frequency
PRF must be larger
than twice the
maximum

17 Figure6.33 Schematic representation of aliasing encountered in Doppler imaging. (a) Same as Figure 6.23. The reflected pulse is
sampled fast enough to avoid aliasing. (b) The scatterer moves faster than in (a) causing a larger phase difference of the subsequent
6 reflected pulses.
Measuring this signal at the range gate tR results in exactly the same sampled sinusoidal function and Doppler frequency fD as in (a).
Sampling is clearly too slow and the calculated scatterer velocity is too low.
 Chapter 6: Ultrasound imaging

spatial frequency c
present in the signal, PRF.
that is, (6.40)
4 va From Eqs. (6.38) and
PRF > 2|f (6.40), it follows that
| |fT.
(6.38) 4|va| c
c fT
<
Because the reflected PRF
ultrasonic pulse must <
be received before the
next one can be (6.4
transmitted, a pulse 1) c
must be able to travel 2d0
to and from the range
gate within the period or
TPRF. Hence, the
c2
following relation
|va| <
between TPRF and the
. (6.42)
depth d0 of the range
gate holds: 8d0fT

2
d0
Equipment
An ultrasound scanner
< is small and mobile.
This is an important
advantage as
T
compared with the
P
other imaging
R
modalities. It consists
F
of a transducer
.
connected to a signal
processing box, which
(
displays the
6
reconstructed images
.
on a monitor in real
3 time (Figure6.34).
9 Laptopandpocket-
) sizeultrasoundscanner
c s are also being
developed. The wide
Because TPRF = 1/PRF applicability of these
by definition, it portable devices will
follows that certainly play an
important role in the
17
7
 Chapter 6: Ultrasound imaging

further development
of ultrasound imaging.
A variety of
transducers for 2D and
3D imaging exist for
different applications.
The most important
characteristics are
discussed below.

One-
dimensiona
l array
transducers
A disadvantage of the
acquisition method
shown in Figure 6.17
is that it involves a
mechanical
displacement or
rotation of the
transducer. This yields
several practical
difficulties. For
Figure6.34 Example of a
example, the commercial
displacement between echocardiographic scanner.
two steps must be (Courtesy of the
Department of Cardiology.)
kept constant, and the
contact with the
medium must not be
lost. To reduce this mechanical
motion, electronic
scanning with an
array
transducercanbeus
ed.
Anarraytransduceri
sacollectionofman
ysmallidenticalcry
stalsthatcanbeexcit
ed independently.
Most common in
medical imaging
are the linear-
17
8
 Chapter 6: Ultrasound imaging

array and the

phased-array
transducers that
consist of a 1D
array of crystals
(Figure 6.35).
The linear-
array transducer
moves the (a) (b)
ultrasonic beam
linearly by firing
its elements
sequentially. This
is the electronic
representation of
the mechanical
translation shown
in Figure 6.17(a).
Typically, linear
arrays are used
where the acoustic
window is large,
which is, for
example, often the
case in vascular
imaging and
obstetrics.
For
applications in
which the acoustic
window is small,
such as in
cardiology where
the waves can
approach the heart
only through the
small space in
between the ribs,
phased-array
transducers can be
used. The crystals
are then used to
steer the direction
of propagation of
the wave by tuning
the phases of the

17
9
 Chapter 6: Ultrasound imaging

Figure6.35 (a) Linear-array transducer.

(b) Phased-array transducer. (Courtesy
of
Professor R. Oyen, Department of

Radiology.)

(a) (b)
Figure6.36 By using an
array of crystals, a sound
beam can be directed (a) or
focused (b) without
mechanical movement of
the transducer. A time
delay is imposed on the
electrical signals that excite
the crystals, causing a
phase shift of the emitted
pulse. This yields the
specific tilted or focusing
wavefront as shown in this
diagram.

waves sent by the

different crystals
(Figure 6.36(a)). This
is the electronic
equivalent of tilting
the transducer
mechanically, as
shown in Figure
6.17(b). At the same
time, the 1D array of
crystals enables the
beam to be focused at
18
0
 Chapter 6: Ultrasound imaging

an arbitrary point phasedarray

along each devices typically
direction (Figure use a constant
6.36(b)). This way sound velocity of
the ultrasonic 1540 m/s. The
beam can be consequence is a
steered and loss of contrast
focused and a and spatial
sector image is resolution. Several
obtained. methods have been
Electronic proposed for phase
focusing is also aberration
useful for linear correction, but
arrays. This can be they are not yet
obtained by firing used in clinical
multiple practice.
neighboring Theoretically the
elements focus is a single point
simultaneously at a certain depth.
(with phase shift) Consequently, an
for each ultrasonic optimal resolution
scan line. Because requires many pulses
the distance along the same scan
between crystals is line, each focusing at
relatively large, a different depth. In
electronic practice, the axial scan
focusing improves line is subdivided into
the resolution. a few segments
The wavefronts around a limited
as shown in Figure number of focal
6.36 are drawn for points. The smaller
the ideal case of an the segments, the
homogeneous better the spatial
medium. For resolution, however,
inhomogeneous at the cost of the
tissue and in the temporal resolution.
case of multiple Small segments are
tissues the velocity unacceptable in
of the ultrasonic cardiac applications,
waves is not but for static organs,
constant and the such as the liver, this
wavefront method is very useful.
becomes distorted. In principle, it is
This effect is not necessary to
called phase transmit multiple
aberration. pulses over time in
Commercial order to change the
18
1
 Chapter 6: Ultrasound imaging

focal point. Indeed, and adding the signals

the individual signals of the individual
received by the array crystals can be used to
of crystals can be change the direction
virtually shifted over of the beam
time after their artificially. Different
reception and before scan lines can then be
they are added obtained
together. simultaneously. This
Consequently, their way the temporal
phase is artificially resolution can be
modified, and the increased from 30 to
focal point during 70–80 Hz, which is
reception changes. If the standard frame
they are changed rate of current clinical
dynamically as a scanners. Higher scan
function of time, the rates require the
focal point is acquisition of
artificially positioned additional scan lines
at the depth from lying further away
which reflections are from the symmetry
expected to arrive axis of the emitted
(because the sound beam, where the beam
velocity in tissue is intensity falls off and,
known). This consequently, the
technique is called detected SNR is
dynamic focusing. It limited.
improves the lateral
resolution
significantly without
influencing the
temporal resolution.
The same principle
of virtually shifting

Rotating transducer Figure6.37 One way of acquiring 3D

Wobbling transducer

datasets is by rotation (a) or wobbling

(b) of the transducer, resulting in a
sampled cone. By reslicing the 3D
dataset, C-mode images are obtained.
C-scans

C-scans

18
2

(a) (b)
 Chapter 6: Ultrasound imaging

the number of 2D
Transducers slices that can be
for 3D acquired per second is
limited by the frame
imaging rate, which in turn is
The easiest way to limited by the sound
construct a 3D image velocity. This way a
is to rotate or wobble 3D acquisition cannot
a phased-array be accomplished in
transducer and acquire real time. Hence, a
images sequentially compromise needs to
from different scan be made between the
planes as illustrated in spatial resolution and
Figure 6.37. Three- the temporal
dimensional imaging resolutions. Moreover,
offers the because a 2D array of
opportunitytoobtainsli crystals contains about
cesofanyorientationthr 64 × 64 = 4096
ough the scanned crystals, A/D
volume. Images of conversion and
planes parallel to the cabling are technically
surface of the difficult.
transducer, as
illustrated in Figure
6.37, are known as C-
scans. Obviously, the
patient should not
move during the
scanning process. An
exception is the cyclic
rhythm of the cardiac
contraction, which can
trigger the imaging
process.
Again, the
mechanical motion Figure6.38 The shape of
the transducer is adapted
can be replaced by its to the application: (1)
electronic abdominal transducers –
general purpose; (2)
representation using a intraoperative
2D phasedarray transducers; (3) small
transducer. The focal parts transducers
(muscles, tendons, skin,
point can be thyroid, breast, scrotum);
positioned at any point (4) intrarectal transducer
(rectal wall, prostate); (5)
within a cone instead intravaginal transducer
of a plane and a 3D (uterus, ovaries,
image can be pregnancy); (6) infants
(abdominal, brain).
obtained. However, (Courtesy of Professor R.
18
3
 Chapter 6: Ultrasound imaging

Oyen, Department of
Radiology.) Clinical
use
Echography is a
Nevertheless,
safe, transportable,
commercial 3D
and relatively
scanners are already
cheap imaging
available. technique that
does not require
any special
Special infrastructure. In
addition,
purpose sequences of
images are
transducers obtained in real
Several time. For these
transducers exist reasons,
for a variety of ultrasoundimaging
applications, such isusuallythemetho
as intrarectal, dofchoiceifit is
intravaginal, clinically relevant.
transesophageal It is useful if the
and intravascular ultrasonic waves
transducers (see are able to reach
Figure 6.38). the tissues under
Transesophageal examination and if
probes (Figure the specular or
6.39) are used to scatter reflections,
visualize the heart or both, are high
through the enough to be
esophagus. perceived in the
Intravascular image.
probes are inserted Consequently,
into an artery or
vein to make
intravascular
ultrasound (IVUS)
images. The
crystal is mounted
on top of a catheter
(diameter 1 mm).
Complete phased
arrays can be built
on this small tip.

Figure6.39
Transesophageal
18
4
 Chapter 6: Ultrasound imaging

transducer for cardiac used (Figure 6.41).

imaging. This
transducer is The
swallowed by the generaldrawbackof
patient and makes suchhigh-
ultrasound images of
the heart from within frequencywavesist
the esophagus. heir limited depth
(Courtesy of the
Department of penetration
Cardiology.) because of their
this method is limited high attenuation
to soft tissues, fluids, coefficient.
and small However, for
calcifications that are tissues at the
preferably close to the surface of the
patient’s body surface body, such as the
and not hidden by retina, this loss of
bony structures. acoustic energy is
negligible.
• Neck The soft
Gray scale tissues close to the
surface, such as
imaging the thyroid (Figure
The most 6.42), salivary
common glands, and lymph
nodes, can easily
investigations be approached by
include the the ultrasonic
following. waves.
• Thorax Air (lungs)
• Head Although and bone (ribs)
ultrasound does limit echography
not completely of the thorax to
penetratethroughb soft tissues that
one, surround the lungs,
thebrainofanewbor such as the pleura
ncan be visualized (Figure 6.43) and
(Figure 6.40) the diaphragm.
because a neonatal Breast
•
skull Mammography
isimmatureandnotf (Figure 6.44) is
ullyossifiedatthefo mostly combined
ntanelles. To with an ultrasound
obtain images of examination for
the retina, high- differential
resolution 20 MHz diagnosis.
transducers are

18
5
 Chapter 6: Ultrasound imaging

(a) (b)

Figure6.40 (a) Normal cranial ultrasound. (b) Fluid-filled

cerebral cavities on both sides as a result of an
intraventricular hemorrhage.
(Courtesy of Professor M. H. Smet, Department of
Radiology.)

(a)

Left AVERAGE
TL = 22.79mm T+12.3dB

18
6
 Chapter 6: Ultrasound imaging

COR=0.00mm AC=3.22mm L=4.69mm V=15.05mm TL= 22.95mm

1620m/s 1 532m/s 1 641m/s 1 532m/s

100%

t(µs)

0 10 20 30
(b)

Figure6.41 (a) B-scans of the eye showing a malignant

melanoma (arrow) with a localized retinal detachment (arrow
head) in a 10 MHz (left) and 20 MHz (right) scan. (b) A-scan of
a patient with cataract (arrow). The dotted lines are
automatically recognized as the borders of the
anterior chamber (AC), the lens (L), and the vitreous (V). The
system calculates the length of each of these regions, taking
the estimated sound velocity in the different tissues into
account. The total eye length is defined as TLare due to
reverberations. (Courtesy of Dr. J. Blanckaert, Department of
Ophthalmology.)= AC + L + V. The small peaks (arrow heads)
in the vitreous

• Abdomen In the diagnosis of

abdominal region, suspicious lesions
all organs such as can be proved by
the spleen, ultrasound-guided
pancreas, and liver biopsy (Figure
(Figure 6.45) are 6.46(b)). The
well suited to be visualization of the
visualized by uterus and ovaries
ultrasound can be optimized
imaging. by transvaginal
scanning.
• Urogenital tract
(The kidney, • Fetus Ultrasound
bladder, prostate, imaging of
testicles, uterus, pregnant women
vagina, and has become daily
ovaries.) Quite routine and
often a transrectal includes an
transducer is used investigation of the
to visualize the fetus (Figure 6.47),
prostate (Figure the uterus, and the
6.46(a)). The placenta.
18
7
 Chapter 6: Ultrasound imaging

• Vascular system purposes but also

Dilations as a guidance
(aneurysm, Figure during
6.48) and interventions, such
obstructions as a biopsy of the
(stenosis, kidney, the
thrombosis) can be prostate or the
recognized in gray liver, and abscess
scale images. drainage.
Often Doppler
imaging yields
complementary Doppler
information (the
clinical use of imaging
Doppler imaging Besides static
will be discussed tissues, motion can
in the next be visualized and
section). velocities
• Musculoskeletal calculated by
means of Doppler
system
imaging.
Echography can be
used to diagnose
tears (Figure 6.49),
calcifications, and
acute
inflammations
with edema in the
muscles,
tendinous, and
capsular structures.
• Heart Cardiac Figure6.42 Ultrasound
ultrasound image of the thyroid
includes images of showing a mild
bilateral enlargement
the ventricles and (arrows) suggesting an
atria (Figure 6.50), inflammatory disease
or hormonal inbalance.
the aorta, the (Courtesy of Dr. D.
valves, and the Bielen, Department of
myocardium. Radiology.)
• Flow imaging
Note that Obviously,
ultrasound Doppler imaging is
imaging is not quite useful for
only used for measuring flow in
diagnostic the blood vessels
18
8
 Chapter 6: Ultrasound imaging

and the heart. estimated in real

Vascular diseases time. The strain
such as (i.e., the local
atherosclerosis and deformation) of
clots (thrombi) in the tissue can then
the big vessels be calculated as
induce local flow the integral of the
disturbances strain rate over
(turbulence, time (Figure 6.52).
velocity changes)
of the blood
because of the
local stricture of Contrast
the vessel. In the
heart, leaking echography
valves (Figure Because the acoustic
6.51) or impedance of air is
interventricular quite different from
shunts can be that of tissue,
demonstrated. ultrasonic waves are
almost completely
• Strainimaging
reflected at a tissue–
This is a more
air interface.
recent application
Consequently, blood
of Doppler
injected with
imaging. When
microscopic air
neighboring pixels
bubbles significantly
move with a
scatters and appears
different velocity,
brighter than normal –
the spatial velocity
even brighter than
gradient can be
tissue (as can be
calculated. This
noticed in Figure
gradient
6.53). Ultrasound
corresponds to the
imaging using a
strain rate (i.e.,
solution with
strain per time
microscopic air
unit; the tissue
bubbles (typical
lengthening or
diameter 4 µm)
shortening per
injected into the blood
time unit). The
circulation is called
strain rate can be

18
9
 Chapter 6: Ultrasound imaging

right
diaphragm kidney
liver

lung effusion

acoustic
shadow of the ribs

(a) (b)

Figure6.43 Ultrasound image of thoracic base: (a) normal

lung, (b) pleural effusion. (Courtesy of Dr. D. Bielen,
Department of Radiology.)
contrast echography. It can be used, for example,
to visualize fluid cavities and for the assessment
of organ perfusion.

Biologic effects and safety

Some experimental animal studies in the late
1960s reported that the genetic properties of
irradiated tissue change. However, these results
have never been confirmed,
andinclinicalpracticethebenefitsofultrasound
imaging outweigh any potential side effects.
Although ultrasound is said to be safe, there are
two physical phenomena that can cause tissue
damage (i.e., tissue heating and cavitation).

Ultrasonic energy is absorbed by the tissue and

19
0
 Chapter 6: Ultrasound imaging

female breast, showing

a hyporeflective lesion
with irregular borders,
which corresponds to a
vast mass with the
characteristics of a
malignant lesion.
(Courtesy of Dr. Van
Ongeval, Department of
Radiology.)
converted to heat.
To prevent tissue
damage, the
thermal index based
on the transmitted
power has been
introduced. This
parameter is
Figure6.44 Ultrasound indicated on
image (10 MHz) of a

(a) (b)

aerobily shadow

(c)

Figure6.45 (a) B-mode image of a normal liver. The liver is

an acoustically homogeneous, large organ. The boundary is
visible as a bright line. The little black holes (arrows) inside
the liver are cross-sections of blood vessels. (b) Liver with
cyst visible as a large black hole (arrow). Note the so-called
acoustic retroamplification, a hyperechoic region behind
the cyst. The origin of this increased reflectivity is the
reduced acoustic attenuation in the fluid within the cyst. (c)
The opposite effect of acoustic retroamplification is
19
1
 Chapter 6: Ultrasound imaging

observed when air is present in the bile ducts, as in a

disease called aerobily. Because air is a perfect reflector for
ultrasound, it is visible as a very bright reflection. Because
of this extremely high attenuation, deeper regions cannot
be imaged and are said to be in the ‘‘acoustic shadow’’ of
the air. (Courtesy of Dr. D. Bielen,
Department of Radiology.)

biopsy probe
prostate

transrectal
transducer

(a) (b)

Figure6.46 Transrectal ultrasound study of the prostate

showing a hypoechoic lesion (arrow) suspicious for
cancer (a). Its nature is proven by ultrasound-guided
biopsy (b). (Courtesy of Professor R. Oyen, Department
of Radiology.)
the ultrasound
scanner and must
not exceed a
certain threshold.
• In areas of low local

density resulting
from a negative
pressure
(rarefaction
regions),
microscopic gas
bubbles can be
Figure6.47 A transverse
ultrasound image of the formed. When the
abdomen of this fetus pressure increases,
shows a fluid-filled and these bubbles
dilated collecting system of
the left kidney (arrows) collapse. This can
resulting from a renal cause tissue
outflow obstruction.
(Courtesy of Dr. I. Witters, damage. A
Department of Obstetrics mechanical index
and Gynecology, and based on the peak
Professor M. H. Smet,
Department of Radiology.) negative pressure
is shown by the
19
2
 Chapter 6: Ultrasound imaging

system and must improvements will

be kept under a contribute to a better
specified image quality, such as
threshold. new silicon based
transducer technology
In diagnostic and real time phase
imaging both effects aberration correction.
are avoided as much Miniaturization of the
as possible. However, transducers and the
they can also be electronics of high-
exploited for therapy. end ultrasound
Heating can be used devices is also a
for ultrasound surgery continuing trend,
making them useful
to burn malignant
for screening and
tissue. Cavitation is
diagnosis outside the
the basis for
practitioner’s office
lithotripsers, which
and for wireless
destroy kidney or
telemedicine.
bladder stones by
Because the
means of high- microstructure of
pressure ultrasound. tissue is defined by the
tissue type (e.g., heart
muscle) and status
(e.g., ischemic or
Future infarcted) and because
this microstructure is
expectation closely related to the
way ultrasonic waves
s are scattered,
From a technical point backscattered signals
of view 3D systems contain information on
with full 2D-array the tissue type, status,
transducers and real or both. Some initial
time 3D velocity and methods of tissue
strain imaging will characterization based
become available. on backscatter have
Increased resolution becomecommerciallya
and contrast-to-noise vailableandsoonerorlat
ratio can be expected erthis technique may
for all imaging be introduced into
modalities. In standard clinical
ultrasound imaging, practice.
several technical
Figure6.48 Sagittal and transversal views of the abdominal
aorta showing an aneurysm (3.1 cm versus a normal size of
1.5–2.0 cm).
(Courtesy of Dr. D. Bielen, Department of Radiology.)

19
3
 Chapter 6: Ultrasound imaging

3.1 cm

(a) (b)

skin

haematoma
deltoid muscle
muscle
fat

supraspinatus m8m
0.
muscle

muscle

humerus

(a) (b)

Figure6.49 (a) Ultrasound image showing a complete tear

of the pectoral muscle with large hematoma, contracted
during a power training session. (Courtesy of Dr. P. Brys,
Department of Radiology.) (b) Ultrasound image showing
an unsharply defined hyperreflective zone because of a
calcification (arrow) and a linear anechoic gap because of
a partial thickness tear of 0.8 mm wide of the
supraspinatus muscle. (Courtesy of Dr. E. Geusens,
Department of Radiology.)

19
4
 Chapter 6: Ultrasound imaging

Figure6.51 Doppler color flow image of a patient

with mitral regurgitation in the left atrium. The bright green
color corresponds Figure6.50 Transesophageal
echocardiographic (TEE) image to high velocities in mixed
directions because of very turbulent flow
showing an atrial septal defect (ASD). (Courtesy of the leaking
through a small hole in the mitral valve. (Courtesy of the
Department of Cardiology.) Department of Cardiology.)

(a) (b)
(c)

Figure6.52 (a) Tissue velocity, represented by color,

obtained by CF imaging. (b) By calculating the velocity
difference between neighboring points, a strain rate
image is obtained. (c) The strain is calculated as the
cumulative strain rate over time. (Courtesy of the
Department of
Cardiology.)

19
5
 Chapter 6: Ultrasound imaging

Figure6.53 B-mode gray scale image of the left ventricle

of the heart in a short-axis view before (a) and during (b)
pervenous injection of a contrast agent.
(Courtesy of the Department of
Cardiology.)

(a) (b)

Further progress yield a local drug

can also be expected delivery.
in contrast echography Very promising is
and its therapeutic the use of
application. In microbubbles with
practice, air quickly targeting ligands that
dissolves in blood, and bind specific
the contrast obtained receptors. This
disappears almost technique, known as
immediately. To targeted contrast-
prevent this, the gas enhanced ultrasound,
can be encapsulated in has a great potential
a thin shell, which for ultrasound
decreases the amount molecular imaging
of scattering but (see Chapter 5, p. 126)
stabilizes the bubbles and targeted drug and
for several hours, gene delivery.
depending on the
material of the gas and
the shell. Recently,
attempts have been
made to load the
encapsulated shells
with drugs with the
intention that the
ultrasonic pressure
will tear the shell and

19
6