J. Anthony Seibert. Physics of Ultrasound PDF
J. Anthony Seibert. Physics of Ultrasound PDF
Physics of Ultrasound
J. Anthony Seibert
Imaging systems using ultrasound have attained a large contraction of the crystal surface by an external power
presence as point-of-care (PoC) devices across many source introduces energy into the medium as a series of
clinical domains over the past 10 years. The success compressions and rarefactions, traveling as a wave front
of ultrasound for this purpose is attributed to several in the direction of travel, known as a longitudinal wave,
characteristics, including the low cost and portability as shown in Fig. 1.1.
of ultrasound devices, the nonionizing nature of ultra-
sound waves, and the ability to produce real-time Wavelength, Frequency, Speed
images of the acoustic properties of the tissues and tis- The wavelength (λ) is the distance between any two
sue structures in the body to deliver timely patient care, repeating points on the wave (a cycle), typically mea-
among many positive attributes. An understanding of sured in millimeters (mm). The frequency (f) is the num-
the basic physics of ultrasound, in addition to hands-on ber of times the wave repeats per second (s), also defined
training, practice, and development of experience are of in hertz (Hz), where 1 Hz = 1 cycle/s. Frequency identi-
great importance in its effective and safe use. This chap- fies the category of sound: less than 15 Hz is infrasound,
ter describes the characteristics, properties, and produc- 15 Hz to 20,000 Hz (20 kHz) is audible sound, and
tion of ultrasound; interaction with tissues, acquisition, above 20 kHz is ultrasound. Medical ultrasound typi-
processing, and display of the ultrasound image; the cally uses frequencies in the million cycles/s megahertz
instrumentation; achievable measurements, including (MHz) range, from 1 to 15 MHz, with some specialized
blood velocity; and safety issues. ultrasound applications beyond 50 MHz. The period is
the time duration of one wave cycle and is equal to 1/f.
The speed of sound, c, is the distance traveled per unit
CHARACTERISTICS OF SOUND time through a medium and is equal to the wavelength
Sound is mechanical energy that propagates through a (distance) divided by the period (time). As frequency is
continuous, elastic medium by the compression (high inversely equal to the period, the product of wavelength
pressure) and rarefaction (low pressure) of particles and frequency is equal to the speed of sound, c = λf. The
that comprise it. Compression is caused by a mechani- speed of sound varies substantially for different mate-
cal inward deformation by an external force, such as an rials, based on compressibility, stiffness, and density
expanding and contracting transducer crystal composed characteristics of the medium. For instance, air is highly
of multiple elements in contact with the medium. During compressible and of low density, with a relatively low
transducer surface expansion, an increase in the local speed of sound; bone is stiff and dense, with a relatively
pressure at contact occurs. Contraction of the crystal very high speed of sound; and soft tissues have com-
follows, causing a decrease in pressure. The mechanical pressibility and density characteristics with intermedi-
energy imparted at the surface is transferred to adjacent ate speeds, as listed in Table 1.1. Of importance are the
particles of the medium, which travels at the speed of average speeds for “soft tissue” (1540 m/s), fatty tissue
sound through the medium. Continuous expansion and (1450 m/s), and air (330 m/s). To relate time with depth
2
Descargado para Anonymous User (n/a) en National Autonomous University of Mexico de ClinicalKey.es por Elsevier en junio 11, 2020.
Para uso personal exclusivamente. No se permiten otros usos sin autorización. Copyright ©2020. Elsevier Inc. Todos los derechos reservados.
CHAPTER 1 Physics of Ultrasound 3
Contraction
Expansion Compression Wavelength, λ
(mm)
Pressure amplitude
1 cycle
Transducer Rarefaction
Fig. 1.1 Mechanical energy is generated from an expanding and contracting crystal in contact with a medium,
introducing high-pressure (compression) and low-pressure (rarefaction) variations of the constituent particles
that transfer the energy to adjacent particles as a longitudinal wave.
Descargado para Anonymous User (n/a) en National Autonomous University of Mexico de ClinicalKey.es por Elsevier en junio 11, 2020.
Para uso personal exclusivamente. No se permiten otros usos sin autorización. Copyright ©2020. Elsevier Inc. Todos los derechos reservados.
4 SECTION 1 Principle of Ultrasound
Descargado para Anonymous User (n/a) en National Autonomous University of Mexico de ClinicalKey.es por Elsevier en junio 11, 2020.
Para uso personal exclusivamente. No se permiten otros usos sin autorización. Copyright ©2020. Elsevier Inc. Todos los derechos reservados.
CHAPTER 1 Physics of Ultrasound 5
θi =θr
θi θr
Reflection Reflection
(echo return to (echo away from
transducer) transducer)
z1 = p1c1
Boundary
z2 = p2c2
Direction Refraction
c2 >c1
unchanged
Transmission θt
c2 <c1
c2 = c1
Fig. 1.3 A boundary separating two tissues with different acoustic impedances demonstrates (A) perpendic-
ular (normal) incidence of an ultrasound wave with reflection of an echo back to the source (transducer) and
transmission to greater depths in a straight line and (B) incidence of the wave at a nonperpendicular angle,
with the incident angle measured relative to the normal incidence and the reflected echo at an angle opposite
but equal to the incident angle. The transmitted ultrasound wave is refracted if the speed of sound is differ-
ent in the two tissues, with the angle of refraction also referenced to the normal direction. Refraction angle
depends on the relative speed differences and the change in the wavelength at the boundary.
Descargado para Anonymous User (n/a) en National Autonomous University of Mexico de ClinicalKey.es por Elsevier en junio 11, 2020.
Para uso personal exclusivamente. No se permiten otros usos sin autorización. Copyright ©2020. Elsevier Inc. Todos los derechos reservados.
6 SECTION 1 Principle of Ultrasound
background signal. Scattered echo signals are more prev- Ultrasound attenuation
alent relative to specular echo signals when using higher µ ≅ 0.5 dB/cm per MHz
1
ultrasound frequencies. 0.9
0.8
Absorption and Attenuation
Relative intensity
0.7
Attenuation is the loss of intensity with distance trav- 0.6
eled, caused by scattering and absorption of the incident 0.5
beam. Scattering has a strong dependence on increasing 0.4
ultrasound frequency. Absorption occurs by transferring 0.3 2 MHz
energy to the tissues that result in heating or mechanical 0.2 5 MHz
disruption of the tissue structure. The combined effects 0.1 10
MHz
of scattering and absorption result in exponential atten- 0
0 510 15 20
uation of ultrasound intensity with distance travelled as Depth of tissue (cm)
a function of increasing frequency. When expressed in Distance traveled = 2 x depth
decibels (dB), a logarithmic measure of intensity, atten- Fig. 1.5 Attenuation and relative intensity of ultrasound remain-
uation in dB/cm linearly increases with ultrasound fre- ing as a function of depth for 2-, 5-, and 10-MHz beams.
quency. An approximate rule of thumb for ultrasound
attenuation average in soft tissue is 0.5 dB/cm times the TABLE 1.2 Attenuation Coefficient μ (dB/
frequency in MHz. Compared with a 1-MHz beam, a cm-MHz) for Tissues*
2-MHz beam will have approximately twice the attenu-
ation, a 5-MHz beam will have five times the attenuation, Tissue μ (1 MHz)
and a 10-MHz beam will have ten times the attenuation Air 1.64
per unit distance traveled. Therefore higher-frequency Blood 0.2
ultrasound beams have a rapidly diminishing penetration Bone 7–10
depth (Fig. 1.5), so careful selection of the transducer fre- Brain 0.6
quency must be made in the context of the imaging depth Cardiac 0.52
needed. The loss of ultrasound intensity in decibels can Connective tissue 1.57
be determined empirically for different tissues by mea- Fat 0.48
suring as a function of distance travelled in centimeters
Liver 0.5
(cm) and is the attenuation coefficient, μ, expressed in
Muscle 1.09
dB/cm. For a given ultrasound frequency, tissues and
Tendon 4.7
fluids have widely varying attenuation coefficients chiefly
resulting from structural and density differences, as indi- Soft tissue (average) 0.54
cated in Table 1.2 for a 1-MHz ultrasound beam. Water 0.0022
*For higher-frequency operation, multiply the attenuation coef-
ficient by the frequency in MHz.
THE ULTRASOUND SYSTEM
PoC ultrasound systems are available from many ven-
dors and come with different features and options, Ultrasound Transducer Operation and Beam
which depend on acquisition capabilities, number of Properties
transducer probes, durability, software functionality, size Ultrasound is produced and detected with a transducer
and weight, battery longevity for handheld units, power array, composed of hundreds of ceramic elements with
requirements, and other considerations. Although all electromechanical (piezoelectric) properties. Ultrasound
ultrasound systems have unique instrumentation, soft- transducers for medical imaging applications employ a
ware, and user interfaces, common components include synthetic piezoelectric ceramic, lead–zirconate–titanate
transducer probes, pulser, beam former, scan converter, (PZT), with a crystal structure that generates a sur-
processor, display, and user interface for instrumenta- face charge of either negative or positive polarity when
tion adjustments and controls. its thickness is expanded under negative pressure or
Descargado para Anonymous User (n/a) en National Autonomous University of Mexico de ClinicalKey.es por Elsevier en junio 11, 2020.
Para uso personal exclusivamente. No se permiten otros usos sin autorización. Copyright ©2020. Elsevier Inc. Todos los derechos reservados.
CHAPTER 1 Physics of Ultrasound 7
Descargado para Anonymous User (n/a) en National Autonomous University of Mexico de ClinicalKey.es por Elsevier en junio 11, 2020.
Para uso personal exclusivamente. No se permiten otros usos sin autorización. Copyright ©2020. Elsevier Inc. Todos los derechos reservados.
8 SECTION 1 Principle of Ultrasound
Transducer wiring
Housing
Absorber
Damping block
Side view
Transducer array
Matching layer
Edge view
Composite structure
Ultrasound beam direction
A
pulse, the process repeats along a slightly different direc- are received from the greatest depths, the next pulse is
tion to ultimately cover the volume of interest defined by created by activating another subelement group that is
the field of view (FOV). incrementally shifted along the transducer array, and
For a given transducer, multifrequency operation the process repeats on the order of thousands of times
allows flexibility of the sonographer to interactively per second to generate a rectangular image format with
choose the appropriate frequency to emphasize the real-time video image capture. Linear arrays are typi-
spatial resolution or depth of penetration based on the cally composed of 256 to 512 transducer elements, are
examination, as shown in Fig. 1.9. of smaller form factor, and generally operate at higher
frequency ranges (5–15 MHz). Because of the higher
Transducer Arrays operating frequency and limited FOV, these transducers
Three basic transducer types for PoC ultrasound are suitable for imaging superficial structures such as the
include linear, curvilinear, and phased arrays, as shown eyes, joints, muscles, and proximal blood vessels and for
in Fig. 1.10. Linear array transducers activate a sub- performing ultrasound-guided procedures. Curvilinear
set of elements, producing a single transmit beam at array transducers have 256 to 512 elements in a convex
one location, and then listen for echoes in the receive geometry, with a subset of elements activated sequen-
mode. Within a fraction of a second, when all echoes tially, like the linear array, producing a trapezoidal
Descargado para Anonymous User (n/a) en National Autonomous University of Mexico de ClinicalKey.es por Elsevier en junio 11, 2020.
Para uso personal exclusivamente. No se permiten otros usos sin autorización. Copyright ©2020. Elsevier Inc. Todos los derechos reservados.
CHAPTER 1 Physics of Ultrasound 9
General
Penetration (Center frequency) Resolution
Selective
transducer
Response
bandwidth
Overall
transducer
bandwidth
4 6 5 7 8 9 10
Frequency (MHz)
Fig. 1.9 Multifrequency transducer transmit and receive response to operational frequency bandwidths
allows the operator to select an appropriate transmit and receive frequency, depending on the type of exam-
ination, type of transducer, transducer bandwidth range, and need for penetration depth (selecting a lower
frequency) or spatial resolution (selecting a higher frequency). The transducer response shown has a select-
able frequency range of 4 to 10 MHz.
image format with increased FOV at both proximal and at a predetermined depth (or depths), which is oper-
distal depths. These transducers are ideal for imaging ator selectable. After excitation, beam direction, and
intraabdominal organs such as the liver, spleen, kidneys, beam formation, the phased array is placed in receive
and bladder. Lower frequencies (2–5 MHz) are used for mode to listen for echoes. The sequence then repeats
depth visualization, but spatial resolution can be lim- along a slightly different beam direction, ultimately
ited as a result. Phased array transducers with 64, 128, creating a sector-shaped image format at a frame rate
and up to 256 elements use all transducer elements in dependent on the number of lines, depth, and FOV. In
the formation of the ultrasound beam. Beam direction the receive mode, returning echoes are detected by all
is determined by relatively large incremental delays active transducer elements used in the formation of the
for sequential excitation of elements from one side of beam. Phased array transducers typically operate at low
the array to the other, effectively steering the beam in frequencies (1–5 MHz), have flexible selection of a nar-
a perpendicular direction to the excitation pattern. row to wide FOV by the operator, and provide efficient
Along a given direction, small incremental excitation two-dimensional (2D) imaging for heart and thoracic
delays in a concave pattern focus the beam diameter imaging requiring small acoustic windows.
Descargado para Anonymous User (n/a) en National Autonomous University of Mexico de ClinicalKey.es por Elsevier en junio 11, 2020.
Para uso personal exclusivamente. No se permiten otros usos sin autorización. Copyright ©2020. Elsevier Inc. Todos los derechos reservados.
10 SECTION 1 Principle of Ultrasound
Spatial Resolution
In ultrasound, the visibility of image detail is determined
by three separate factors: (1) in-plane resolution along
the direction of beam travel, known as axial resolution;
(2) in-plane resolution perpendicular to the direction of
beam travel, known as lateral resolution; and (3) out-of-
plane resolution perpendicular to the in-plane resolution, Lateral
known as elevational or slice-thickness resolution. These
constituents of spatial resolution are illustrated in Fig. 1.12. Elevational
Axial resolution represents the ability to distinctly
separate closely spaced objects in the direction of the
ultrasound beam. Returning echoes from adjacent Axial
boundaries to be resolved as separate are dependent
on the spatial pulse length (SPL), which is the average
wavelength times the number of cycles in the pulse.
Because distance travelled for a pulse-echo interacting Fig. 1.12 The three components of spatial resolution in the
between two adjacent reflectors is twice the separation ultrasound image are shown. Axial, along the beam direction,
is constant with depth. Lateral, in-plane and perpendicular to
distance, echoes to be recorded as separate signals need the beam direction, varies substantially with depth. Elevational,
a spacing just greater than one-half SPL. For example, a perpendicular to the lateral and axial directions, is the slice
5-MHz frequency pulse has a wavelength of 0.31 mm in thickness and varies with depth.
Descargado para Anonymous User (n/a) en National Autonomous University of Mexico de ClinicalKey.es por Elsevier en junio 11, 2020.
Para uso personal exclusivamente. No se permiten otros usos sin autorización. Copyright ©2020. Elsevier Inc. Todos los derechos reservados.
CHAPTER 1 Physics of Ultrasound 11
soft tissue, consisting of three cycles. The SPL is 3 × 0.31 is dependent on the maximum image depth. Between
= 0.93 mm, and the axial resolution is ½(0.93 mm) = pulses is the pulse repetition period (PRP), equal to 1/
0.46 mm. A 10-MHz frequency pulse with three cycles PRF, and is the time that the transducer is in receive
has SPL = 0.46 mm, with axial resolution of ½(0.46) = mode to listen for returning echoes. A 2-kHz PRF has a
0.23 mm. Axial resolution is independent of depth. To PRP = 1/2000 s-1 = 0.0005 s = 0.5 ms. Within this time,
achieve enhanced axial resolution, use of a higher-fre- an ultrasound pulse in tissue can propagate to a depth of
quency transducer operation can be selected, but pene- 38.5 cm and return as an echo to the transducer before
tration depth may be compromised and inadequate. the next pulse. Thus the maximum PRF is determined
Lateral resolution refers to the ability to resolve by the time required for echoes from the most distant
objects of a given size perpendicular to the beam direc- structures to reach the transducer; otherwise, these
tion and is directly dependent on the beam diameter, echoes can be confused with prompt echoes from the
which varies as a function of depth and is best within next pulse, resulting in range ambiguity artifact. Higher
the focal zone. Lateral resolution worsens proximal transducer frequencies with greater attenuation and
and distal to the focal zone. The operator can adjust the limited penetration depth allow higher PRFs. The PRF
lateral focal zone typically to one depth, and in some determines the trade-off of image frame rate (tempo-
higher-end systems at several depths within the image; ral resolution) versus image quality (number of lines/
however, the trade-off is a loss of temporal resolution frame) and is particularly important in Doppler ultra-
or number of scan lines per image. Lateral resolution is sound, as there is a direct effect on the sampling rate to
typically two to three times less than axial resolution. accurately determine the velocity of moving objects (see
Elevational resolution indicates the ability to dis- the section below on Doppler ultrasound).
tinctly resolve acoustically generated objects repre-
sented in the slice-thickness dimension of the image, Hardware Components and Data Processing
which is perpendicular to the displayed image plane. In Acquisition of ultrasound data involves several hard-
this dimension, the beam thickness varies like the lateral ware components: pulser, beam former, transmit/
beam dimension, where close to the transducer the beam receive switch, receivers, amplifiers, and scan converter,
thickness is large, converges over a region at mid-depth, as shown in Fig. 1.13. The pulser (also known as the
and then diverges with continued propagation. Volume
averaging causes objects close to the transducer sur-
face or at greater depths to be not as well depicted. Slice Pulsers Beam former Control
digital steering/focusing panel
thickness is typically the weakest measure of resolution beam summation
for array transducers. The ability to vary elevational res- Transmit/receive
olution is possible with sophisticated transducers, not switches
usually found for PoC ultrasound systems. Receiver
time gain
Pre-amps A/Ds
log compression
ULTRASOUND DATA ACQUISITION rectification
rejection
Swept gain
Data acquisition and image formation occur in a pulse–
echo mode and are operationally dependent on creating
an ultrasound pulse, listening for echoes, and constantly Image
Patient
repeating the procedure over the acquisition event. The Scan converter
ultrasound pulse is intermittently produced by activat- memory
ing transducer elements at a rate known as the pulse rep- digital interface
postprocessing
etition frequency (PRF), equal to the number of pulses storage
produced per second. For data acquisition and imaging,
the PRF typically ranges from 1000 to 5000 pulses/s (1–5 Fig. 1.13 Components of an ultrasound system needed to
acquire ultrasound image data is shown. The grayscale image
kHz), depending on the application, depth of echoes to is created from returning echoes detected as a function of time
be recorded, and frame rate desired, among other param- and ultrasound beam direction from A-mode data, converted to
eters. Usually, the PRF is automatically determined and B-mode, digitized, and mapped into a 2D digital image matrix.
Descargado para Anonymous User (n/a) en National Autonomous University of Mexico de ClinicalKey.es por Elsevier en junio 11, 2020.
Para uso personal exclusivamente. No se permiten otros usos sin autorización. Copyright ©2020. Elsevier Inc. Todos los derechos reservados.
12 SECTION 1 Principle of Ultrasound
Received echoes
Amplitude
Before
TGC
TGC Gain
amplification
Amplitude
After
TGC
A-line:
Amplitude
Compression
Demodulation
Rejection
A Time (depth) B
Fig. 1.14 (A) Time gain compensation (TGC) is a user adjustment to make equally reflective boundaries
appear with the same amplitude, despite attenuation with depth. (B) TGC controls are provided as operator
slider bars on the console, electronic sliders on the video monitor, or other methods that allow image gain
adjustments at a specific depth.
transmitter) provides the electrical voltage for activating crystals are unable to detect proximal echoes, creates
the piezoelectric transducer elements. Transmit power a “dead zone” image region at very shallow depth. For
is adjusted by the excitation voltage on the transducer evaluation of surface or near-surface regions of the body,
crystal—a higher voltage creates a higher-intensity an acoustic stand-off window is often used. The receiver
ultrasound pulse and improves echo detection from accepts data and performs signal processing, including
weak reflectors but increases power deposition to the time gain compensation (TGC), dynamic range com-
patient. Transmit power settings are adjusted for the pression, signal demodulation, and noise rejection.
examination type; for instance, a low power is used for TGC (also described as time-varied gain and depth gain
obstetric imaging. Pulse duration is the instantaneous compensation) is a user-adjustable amplification of the
“on” time of the ultrasound pulse, on the order of 0.001 returning echo signals as a function of time to compen-
ms (one-millionth of a second, or 1 μs). Duty cycle, the sate for ultrasound attenuation. An ideal adjustment of
fraction of the ultrasound beam “on” time, is equal to TGC results in equally reflective boundaries with the
the pulse duration divided by the PRP and is typically same signal amplitude, independent of depth, as shown
0.2% to 0.4% for imaging—thus more than 99.5% of in Fig. 1.14A. Depending on equipment, user adjust-
ultrasound scan time is spent in the receive mode, lis- ment occurs with multiple slider potentiometers desig-
tening for returning echoes. The beam former provides nated for a certain image depth (Fig. 1.14B) or by TGC
active beam steering as well as transmit and receive controllers to manipulate initial gain, slope, and far gain
focusing. The transmit/receive switch isolates the high of the echo signals. Subsequent processing steps include
voltage (≈150 V) applied to the transducer elements signal compression, rectification, demodulation, and
during excitation from extremely low voltages (milli- noise rejection. The output is an amplitude-modulated
volt to microvolt range) induced by returning echoes. At (A-mode) line of data representing a time sequence of
the initial excitation, a ring-down time, when the PZT detected echoes.
Descargado para Anonymous User (n/a) en National Autonomous University of Mexico de ClinicalKey.es por Elsevier en junio 11, 2020.
Para uso personal exclusivamente. No se permiten otros usos sin autorización. Copyright ©2020. Elsevier Inc. Todos los derechos reservados.
CHAPTER 1 Physics of Ultrasound 13
Depth
Returning
echoes
The scan converter, image memory, and display is usually 24 bits (3 bytes), with each byte encoding for
monitor are required to process A-mode line data into a red, green, and blue color combinations, and an uncom-
brightness (B-mode) signal to create time–motion (T-M pressed size of ¾ MB. Image compression schemes such
or M-mode) and grayscale image (B-scan) outputs. as JPEG and MPEG are often used in the final output
B-mode represents the echo amplitudes converted into image or image stream to substantially reduce the over-
proportional brightness-modulated dots as a function of all size of the image data.
depth (time) along the A-line trajectory. M-mode uses a Time for a single ultrasound image is determined
stationary transducer positioned over moving anatomy, by the number (N) of A-lines acquired across the FOV
such as valve leaflets, where motion in the ultrasound (the line density) and the PRF/PRP (governed by the
beam is tracked by location of B-mode brightness dots, needed depth of penetration). These parameters are
recorded as a function of depth (vertical position) and part of acquisition protocols and can be directly or
time (horizontal deflection) of the traces to enable diag- indirectly changed by operator adjustments. Protocols
nosis of periodic or aperiodic motion, as illustrated in represent a decision of diagnostic need in terms of
Fig. 1.15. temporal resolution versus image quality. For instance,
Two-dimensional grayscale images are constructed if high frame rates are needed, the number of A-lines,
by identifying beam directions relative to the transducer FOV, or penetration depth (with high-frequency oper-
position with the scan converter and mapping B-mode ation to allow a higher PRF) can be reduced. When
data to build an image in computer memory for display image quality is more important, line density (num-
on the device monitor. Digital processing is subsequently ber of A-lines sampled across the FOV) is increased,
applied, with application of a variety of electronic and with a loss of frame rate. Insufficient line density can
mathematical functions, including gain, window width cause the image to appear pixelated, which is caused by
and window level adjustment, edge enhancement, noise interpolation of several pixels to fill unscanned image
reduction, and data interpolation for diverging beams, regions. For sector and trapezoidal scans, increased
among other schemes. Each image is typically rendered line spacing with depth also results in decreased image
into a 512 × 512 × 8 bit (1 byte) per pixel matrix, equal quality. Increasing the number of A-lines sampled over
to ¼ megabyte (MB). For color encoding, the bit depth a specified FOV or reducing the FOV for the same
Descargado para Anonymous User (n/a) en National Autonomous University of Mexico de ClinicalKey.es por Elsevier en junio 11, 2020.
Para uso personal exclusivamente. No se permiten otros usos sin autorización. Copyright ©2020. Elsevier Inc. Todos los derechos reservados.
14 SECTION 1 Principle of Ultrasound
Fig. 1.16 Factors determining the trade-off between image quality and temporal resolution are field of view,
penetration depth, pulse repetition frequency/pulse repetition period, and ultrasound line density, as shown.
Examination requirements identify parameters that must be considered when designing an acquisition pro-
tocol.
number of lines can achieve adequate line density. Harmonic imaging is a method that uses higher fre-
These trade-offs are illustrated in Fig. 1.16 for a sector quencies generated by nonlinear propagation of ultra-
scan. On advanced systems, multiple lateral focal zones sound in tissues to introduce harmonics, which are
that can be set by the operator provide better lateral integer multiples (e.g., 2×, 3×) of the incident frequency.
depth resolution, but at the loss of frame rate and tem- A distortion of the wave occurs as the high-pressure
poral resolution. Ultrasound image rates from 10 to compression part of the wave travels faster than the
up to 60 frames per second (and higher for specialized low-pressure rarefaction, introducing higher-order fre-
ultrasound equipment) are typically achieved, with quency harmonics that localize in the central area of the
trade-offs of image quality versus temporal resolution low frequency beam, as shown in Fig. 1.18. Reflected
that must be considered, given the requirements of the echoes of the harmonic frequencies (typically the second
examination. harmonic) are detected by use of a multifrequency trans-
ducer set to the higher frequency. For instance, a transmit
Specialized Acquisition Modes frequency of 2 MHz (3 MHz) has the receive frequency
Spatial compounding is an option on many PoC ultra- tuned to 4 MHz (6 MHz). This reduces low-frequency
sound systems to obtain ultrasound data from several echo clutter occurring in the shallow regions of the
different angles of insonation (typically from 3 to 5), image and benefits from improved axial resolution due
which are subsequently combined to produce a single to higher-frequency returning echoes. As the harmonic
image, as shown in Fig. 1.17. As each image is produced frequencies are concentrated in the central area of the
from data derived from multiple beam angles, the prob- beam, lateral resolution is also improved by the smaller
ability of perpendicular incidence to a boundary reflec- effective beam diameter. Although not always advanta-
tor is increased, providing better boundary definition, geous, native tissue harmonic imaging is best applied in
more continuity with curved structures, and improved exams requiring a lower transmit transducer frequency
signal-to-noise ratio due to data averaging. A downside for depth penetration, allowing for the harmonics to
to spatial compounding is the loss of temporal resolution build and return as a higher-frequency harmonic echo,
frame rate by a factor equal to the number of insonation but without too much attenuation returning to the trans-
angles and the loss of spatial resolution when moving ducer. With the receiver frequency switched to the high-
anatomy is present in the scan. Spatial compounding is er-frequency harmonic, the outcome is improved spatial
less useful for imaging situations that have substantial resolution and substantially less clutter from proximal
voluntary or involuntary patient motion. low-frequency echoes (Fig. 1.19).
Descargado para Anonymous User (n/a) en National Autonomous University of Mexico de ClinicalKey.es por Elsevier en junio 11, 2020.
Para uso personal exclusivamente. No se permiten otros usos sin autorización. Copyright ©2020. Elsevier Inc. Todos los derechos reservados.
CHAPTER 1 Physics of Ultrasound 15
Fig. 1.17 Spatial compounding is applied to linear, curvilinear, and phased array transducer acquisitions by
transmitting ultrasound into the tissues at slightly different angles to increase reception of returning echoes
from otherwise nonperpendicular boundaries and decreasing image noise by averaging the results.
Slower
Increasing depth of tissue
Ultrasound Image Modifications enhance the image to delineate details within the image
Grayscale ultrasound images can be modified by window that are otherwise blurred. Two methods, “read” zoom
width and window level adjustments at the control panel and “write” zoom, are usually available. “Read” zoom
to nondestructively modify the brightness and contrast enlarges a user-defined region of the stored image and
of the displayed image, implemented with l ook-up-table expands the information over a larger number of pixels
(LUT) transformations. The pixel density can limit in the displayed image with replication and interpola-
the quality and resolution of the displayed image. A tion. Even though the displayed region becomes larger,
“zoom” feature on many ultrasound instruments can the resolution of the image itself does not change. Using
Descargado para Anonymous User (n/a) en National Autonomous University of Mexico de ClinicalKey.es por Elsevier en junio 11, 2020.
Para uso personal exclusivamente. No se permiten otros usos sin autorización. Copyright ©2020. Elsevier Inc. Todos los derechos reservados.
16 SECTION 1 Principle of Ultrasound
Transducer bandwidth
Transmit Receive
ƒo Frequency 2ƒo
Normal
Harmonic
Fig. 1.19 Harmonic imaging is achieved with a multifrequency transducer by using a low-frequency transmit
pulse (e.g., 2 MHz) and a higher-order harmonic receive frequency (e.g., 4 MHz). Normal and harmonic images
depict the noticeable improvement in image contrast and resolution.
“write” zoom requires the operator to rescan the area of over equipment parameters such as selection of trans-
the patient that corresponds to the user-selectable area. ducer frequency, ultrasound intensity through transmit
When enabled, the transducer scans the selected area; gain settings, TGC curves, and noise threshold settings,
echo data within the limited region are acquired; and among others. Measures of image quality include spatial
line density is increased to improve image data sam- resolution, contrast resolution, image uniformity, and
pling, spatial resolution, and contrast. noise characteristics. Additionally, image artifacts can
Besides the B-mode data used for the 2D image, other enhance or degrade the diagnostic value of the ultra-
information from M-mode and Doppler signal pro- sound image, as covered in Chapter 2, on ultrasound
cessing (to be discussed) can also be displayed. During artifacts. Ultrasound spatial resolution, as previously
operation of the ultrasound scanner, information in the described, includes axial, lateral, and elevational res-
memory is continuously updated in real time. When olutions. Axial and lateral resolutions are in the plane
ultrasound scanning is stopped, the last image acquired of the image. Elevational (slice-thickness) resolution,
is displayed on the screen until ultrasound scanning perpendicular to the plane of the image, is not directly
resumes. discernable. It varies as a function of depth, like lat-
eral resolution, but is typically fixed and not adjustable.
Image Quality Contrast resolution is the ability to discern differences
Image quality is dependent on the ultrasound equip- in acoustic impedance that are assigned certain gray-
ment, transducers, frequency, modes of imaging scale values as rendered on the display monitor. Degra-
selected, positioning skills of the operator, and acqui- dations can limit contrast, including insufficient signal
sition protocols, which should be set with examina- strength with transmit gain set too low, very large
tion-specific needs in mind. The operator has control patients, inappropriate TGC adjustments, anatomic
Descargado para Anonymous User (n/a) en National Autonomous University of Mexico de ClinicalKey.es por Elsevier en junio 11, 2020.
Para uso personal exclusivamente. No se permiten otros usos sin autorización. Copyright ©2020. Elsevier Inc. Todos los derechos reservados.
CHAPTER 1 Physics of Ultrasound 17
Fig. 1.20 Measurement of the biparietal diameter of a developing fetus, indicating 7.82 cm measurement
and gestational age of 31 weeks and 3 days. Notice the caliper markers along the axial direction of the beam.
Also note the dropout of the curved surface of the skull due to echoes reflecting away from the transducer.
clutter, excessive electronic noise, transducer element ultrasound beam are usually more reliable because of
malfunction, image artifacts, display monitor quality/ better axial spatial resolution. Measurement accuracy
calibration, and room viewing conditions, many of is improved by careful selection of reference positions,
which the operator has no control over. Image noise is such as the leading edge of the first boundary to the
chiefly generated by electronic amplification of weaker leading edge of the second boundary along the axis of
signals with depth, so deeper anatomic regions have the beam. Along the lateral direction, the echoes are
diminished contrast-to-noise ratios. Image processing smeared out due to poorer lateral resolution, depen-
that specifically reduces noise, such as temporal or spa- dent on the beam diameter at a given depth, with less
tial averaging, can increase the contrast-to-noise ratio; reproducibility and accuracy. Distance measurements
however, trade-offs include lower frame rates and/or can be extended in a straightforward way to area mea-
poorer spatial resolution. Spatial compounding better surements by assuming a specific geometric shape in
depicts tissue boundaries with multiple-angle inci- the plane of the image. With multiple planar images,
dence of the ultrasound beam and reduces stochastic area measurements can likewise extend to three-di-
speckle and electronic noise through averaging. Har- mensional (3D) volumes by estimating the slice thick-
monic imaging improves image contrast by reducing ness as a function of depth. An example is illustrated in
clutter due to low-frequency echoes. Although these Fig. 1.20 of an obstetric examination and the measure-
and other tools and acquisition modes can assist in ment of biparietal diameter to estimate gestational age
delivering the best image quality possible, it is also very based on the age-related values. Ultrasound assumes a
important to recognize the limitations that can only be speed of 1540 m/s to determine distances and is very
partially mitigated. accurate in these types of examinations. When the
speed of sound is largely different (e.g., in fat and bone)
Quantitative Ultrasound Measurements the accuracy of measurements can be compromised, so
Distance, area, and volume measurements are rou- acknowledgment of these issues is important.
tinely and accurately performed on calibrated diag-
nostic ultrasound systems, based on the relationship of
ultrasound speed to the round-trip time of the pulse
DOPPLER ULTRASOUND
and the echo. From the physics perspective, mea- Doppler ultrasound assesses the velocity of moving
surements between points along the direction of the reflectors, typically blood cells in the vasculature,
Descargado para Anonymous User (n/a) en National Autonomous University of Mexico de ClinicalKey.es por Elsevier en junio 11, 2020.
Para uso personal exclusivamente. No se permiten otros usos sin autorización. Copyright ©2020. Elsevier Inc. Todos los derechos reservados.
18 SECTION 1 Principle of Ultrasound
where v is the velocity of the blood cells and c is the a difference in the observed changes in wavelength and
velocity of ultrasound. The factor of 2 arises to com- therefore frequency, which must be corrected for accu-
pensate for transmitting the ultrasound and receiving a rate evaluation of the true blood velocity.
reflection. As an example, blood cell velocity of 30 cm/s To quantitatively measure blood velocity, a “duplex”
is a tiny fraction of the speed of sound, equal to 154,000 mode of operation allows the user to simultaneously
cm/s. For an incident frequency of 4 MHz, the Doppler perform grayscale B-mode imaging and pulsed Dop-
shift at this specific measurement point in time is cal- pler evaluation, the latter with a separate transducer
culated as element group within the transducer array. Over a
30 vessel of interest visualized in the image, a user-posi-
× 4 × 106 Hz × 2 ≅ 1558hz ≅ 1.5kHz . tioned gate region is located and a Doppler angle, θ,
154, 000
relative to the vessel is determined, as shown in Fig.
One can hear the frequency shift on an audio speaker, 1.22. For blood moving toward the transducer, the fre-
as this is within the audible frequency range. With a quency that is proportional to the velocity vector along
periodic heartbeat and pulsatile blood motion, velocity the blood vessel axis is greater than the frequency mea-
changes can be heard as changes in pitch of the Doppler sured along the direction of the Doppler transducer,
frequency. If the observation and returning echoes are at causing the measured Doppler shift, fd , to be less by
an angle relative to the direction of motion, there will be a fraction equal to the cosine of the Doppler angle as
Descargado para Anonymous User (n/a) en National Autonomous University of Mexico de ClinicalKey.es por Elsevier en junio 11, 2020.
Para uso personal exclusivamente. No se permiten otros usos sin autorización. Copyright ©2020. Elsevier Inc. Todos los derechos reservados.
CHAPTER 1 Physics of Ultrasound 19
Doppler angle θ
Measured
velocity
at angle θ
Doppler gate
Velocity at
angle θ
Fig. 1.22 (Left) Doppler evaluation of frequency shift associated with blood velocity, illustrating the Doppler
gate, the Doppler angle, and the calculated velocity vectors in the direction of blood flow and in the direction
of the transducer. (Right) Image acquired with duplex acquisition and color flow active area (trapezoidal white
boundary, see the text for an explanation) shows the vessel location that aids in placing the Doppler gate.
determined by trigonometry. Thus cos(θ) is the correc- cos 63° = 0.454, cos 80° = 0.174, and cos 84° = 0.105.
tion to the calculated Doppler shift at an angle other Plugging these values into the equation results in a
than 0 degrees, resulting in the generalized Doppler velocity estimate error of ≈10% for a 63-degree Doppler
shift equation: angle (actual 60 degrees) and ≈66% for an 84-degree
2fi v cos(θ) Doppler angle (actual 80 degrees). Doppler angles less
fd = than 20 degrees are also problematic because of the dif-
c
ficulty in clearly identifying the vessel in profile, and
Blood velocity is determined by rearranging the with pulsed Doppler methods (discussed later), the
equation and solving for v: measurements are prone to insufficient sampling rates
fd c to identify the maximum Doppler shift. High-velocity
v= blood flow can exhibit an artifact known as aliasing,
2fi cos (θ) which is represented as reverse velocity (flow) in the
quantitative estimate.
where the measured Doppler shift is adjusted by 1/
cos(θ). As the Doppler angle increases, the measured Continuous and Pulsed Doppler Modes
Doppler shift decreases, necessitating correction to Two distinct methods using continuous wave or pulsed
determine the actual blood velocity. Typical Doppler ultrasound transducer operation are used to mea-
angles range from 30 to 60 degrees for two major rea- sure blood velocity. The continuous wave Doppler
sons: (1) the vessel should be displayed in profile, which system is the simplest and least expensive, with two
needs a minimum of about 30 degrees relative to the transducers—a transmitter for the incident ultrasound
vessel axis for recognition in the image, and (2) at angles generating continuous, narrow-bandwidth ultrasound
greater than 60 degrees, small errors in the angle esti- and a receiver detecting the returning continuous
mate cause increasingly larger errors in the velocity cor- echoes from the vasculature. Positional adjustment of
rection factor 1 / cos(θ) as the Doppler angle approaches the transmit and receive transducers identifies an over-
90 degrees, and thus large errors in the actual veloc- lap area that includes the vessel of interest. In opera-
ity. For example, making a +5% error in the Doppler tion, a demodulator compares the transmit and receive
angle at 60 degrees is 63 degrees and at 80 degrees is 84 frequencies from which the Doppler shift frequency is
degrees. Corresponding cosine values are cos 60° = 0.5, extracted. An amplifier converts the Doppler shift signal
Descargado para Anonymous User (n/a) en National Autonomous University of Mexico de ClinicalKey.es por Elsevier en junio 11, 2020.
Para uso personal exclusivamente. No se permiten otros usos sin autorización. Copyright ©2020. Elsevier Inc. Todos los derechos reservados.
20 SECTION 1 Principle of Ultrasound
into an audio output, and a recorder tracks spectrum shows the relationship of the maximum blood
changes as a function of time for analysis of pulsatile velocity to the Doppler PRF, Doppler angle, and trans-
flow. The advantages of continuous mode include high ducer frequency. To obtain an accurate measurement
accuracy of the Doppler shift measurement due to a of a high velocity, Doppler PRF can be increased, inci-
narrow operational frequency bandwidth and no signal dent transducer frequency can be decreased, or the
aliasing of high-velocity motion because of continu- Doppler angle can be increased. Of these adjustments,
ous transmit and receive operation. The disadvantages the most interactive is adjustment of the Doppler angle
include difficulty in determining depth and overlap area by the operator. If a 1.6-kHz maximum Doppler shift
from which to evaluate a vessel, multiple overlying ves- is present in the measured signal, a PRF of 2 × 1.6 kHz
sels making it difficult to distinguish a specific Doppler = 3.2 kHz is needed to avoid aliasing. The Doppler
signal, and moving objects within either the transmit PRF cannot be set to arbitrarily high values because
or receive beams that can disrupt frequency shifts and of the ultrasound transit time required during the
cause errors in the velocity estimates. PRP. A larger Doppler angle (e.g., 60 degrees) reduces
Pulsed Doppler ultrasound is used in conjunction the Doppler shift frequencies in the measured signal,
with B-mode and color flow imaging to identify vessels which are subsequently adjusted to determine the true
of interest. A separate transducer element array is used blood velocity. At angles larger than 60 degrees, how-
for Doppler operation, with a lighter damping block ever, small errors in angle estimation cause significant
function providing a narrow-frequency bandwidth and errors in the estimation of blood velocity, as explained
longer SPL (see Fig. 1.8). With interactive panel con- previously.
trols, the user positions and adjusts a region and size Blood flow (in units of cm3/s or milliliters/s) is esti-
over the vessel to be interrogated. A Doppler angle is mated as the product of the vessel’s cross-sectional area
estimated from the direction of the Doppler transmit– (cm2) times the velocity (cm/s). Errors in flow volume
receive pulse relative to the vessel axis. Electronic logic may occur due to several circumstances. The vessel axis
and timing algorithms reject all echo signals except those might not lie totally within the scanned plane, or flow
falling within the gate region. Each interrogation pulse might be altered from the perceived direction. The Dop-
and returning echo represent a discrete sample of the pler gate (sample area) could be mispositioned or of
frequency shift caused by reflections of moving blood inappropriate size, such that the velocities are an overes-
cells within the region, measured as a phase change. The timate (gate area too small) or underestimate (gate area
Doppler transducer PRF represents the sampling rate too large) of the average velocity. In addition, errors in
of the returning echo signals and their frequencies. The the cross-sectional area evaluation will cause errors in
maximum frequency accurately determined is governed the flow estimate.
by the Nyquist sampling requirement, where a frequency
signal requires at least two samples per cycle, as shown in Doppler Spectrum
Fig. 1.23. Thus the maximum frequency shift accurately The Doppler spectrum is a display of the Doppler shift
measured is one-half of the PRF. When Doppler shift fre- frequencies contained in the Doppler gate region as a
quencies exist beyond this maximum, such as stenotic function of time, as illustrated in Fig. 1.24. This infor-
jets with very high blood velocity, artifactual signals are mation is displayed on the ultrasound monitor below
generated due to insufficient sampling and are repre- the B-mode image as a moving trace, with a range of
sented as reverse flow. In a pulsed Doppler acquisition, blood velocities measured from −Vmax to +Vmax on the
the maximum Doppler frequency shift, fd,max, is equated vertical axis and time on the horizontal axis. The spec-
to PRF / 2, and the Doppler shift equation becomes trum appears as a periodic waveform with a narrow or
PRF 2fi vmax cos (θ) . filled-in trace, depending on the range of frequencies
Δfd , max = = contained in the signals analyzed from the Doppler gate
2 c
region, and the wall filter settings that remove low-fre-
Rearranging the equation and solving for vmax in terms quency motion not associated with blood flow. As new
of PRF, data arrive, the information is updated and scrolled
c × PRF
vmax = from left to right. Pulsatile blood takes on the appear-
4fi cos (θ) ance of a choppy repeating wave through the periodic
Descargado para Anonymous User (n/a) en National Autonomous University of Mexico de ClinicalKey.es por Elsevier en junio 11, 2020.
Para uso personal exclusivamente. No se permiten otros usos sin autorización. Copyright ©2020. Elsevier Inc. Todos los derechos reservados.
CHAPTER 1 Physics of Ultrasound 21
Pulse Echo
Phase change
Wall filter
> 2 samples/cycle
Doppler shift
frequency
High-frequency
Doppler shift
< 2 samples/cycle
Assigned (aliased)
frequency Time
Fig. 1.23 Pulsed Doppler operation evaluates the phase change between the incident pulse and the returning
echo that is mapped for each sample and recorded as a phase change. A wall filter smooths the resultant
frequencies contained in the evaluation to create the constituent Doppler frequencies. If higher frequencies
caused by high blood velocity generate Doppler frequencies greater than one-half the Doppler pulse repetition
frequency (two samples/cycle), the assigned frequency values will be misrepresented as lower-frequency
signals of opposite phase and will be aliased as slow velocity in the reverse direction (bottom).
Doppler spectrum
avg 0
0
min
–0.2 m/s –0.4 m/s
Time
aliasing
Adjusted baseline Symmetrical baseline
Fig. 1.24 The Doppler spectrum is a plot of the Doppler frequencies present in the Doppler gate region,
plotted as a function of time. Scaling of the range of velocities is dependent on the PRF of the Doppler trans-
ducer elements; in this example, symmetrical allocation is shown on the spectrum, but aliasing is evident
(right). To avoid aliasing, the operator can adjust the baseline to reallocate sampling (left). Maximum, average,
and minimum velocities are extracted to calculate pertinent clinical measures, including pulsatility index and
resistive index.
Descargado para Anonymous User (n/a) en National Autonomous University of Mexico de ClinicalKey.es por Elsevier en junio 11, 2020.
Para uso personal exclusivamente. No se permiten otros usos sin autorización. Copyright ©2020. Elsevier Inc. Todos los derechos reservados.
22 SECTION 1 Principle of Ultrasound
cycle of the heartbeat. Depending on the Doppler gate Color Flow Imaging
positioning, vessel size, and shape, blood movement Color flow imaging (also known as color Doppler) is
exhibits laminar, blunt, or turbulent flow patterns. In a 2D visual display of moving blood superimposed
the center of large vessels, fast and laminar flow occurs; on the conventional grayscale image, as shown in Fig.
near vessel walls, frictional forces slow blood velocity. 1.25. In this mode, a subarea of the image is activated,
Plaque buildup on vessel walls and narrowing (stenosis) with multiple small subregions individually evalu-
of the vessel result in turbulent flow. When a Doppler ated to identify areas of motion using phase-shift or
gate area is positioned to encompass the entire vessel time-domain autocorrelation techniques. These calcu-
lumen, there exists a large range of blood velocities and lation methods are very fast to enable color-encoded,
corresponding large range of Doppler frequencies in real-time updates of blood vessel velocity that are
the spectrum. Conversely, a centrally positioned small
gate contains a narrow range of faster velocities and nar-
row-frequency content. A Doppler gate positioned over
a stenosis results in the largest range of velocities and Timing and digital logic
the greatest likelihood of aliasing. The spectral Doppler
display demonstrates the presence, direction, and veloc- Scan
ity characteristics of blood movement. The direction of Transducer Pulse-ECHO converter
ARRAY and imaging
flow is best determined with a small Doppler angle of beam former system Image
formatter
about 30 degrees. Normal versus diseased vessels can be
characterized by the respective spectral Doppler display
waveforms and shapes due to their hemodynamic fea-
tures. Quantitative vascular measures such as pulsatility
index (PI) and resistive index (RI) are extracted from
the Doppler spectrum waveform using the maximum,
minimum, and average velocity values. PI = (max − Color-flow Evaluation
min) / average, and RI = (max − min) / max, which can “active” area samples
be very useful in many clinical situations.
In a spectral Doppler display, aliased signals demon-
strate reversed flow with negative amplitudes. Reduc-
ing or eliminating aliasing errors requires the user to Grayscale
adjust the spectral Doppler velocity scale to a wider “active” area
range, as the PRF of the Doppler element is linked to
the scale setting. When the maximum PRF has been
reached, the spectral baseline, which represents 0 veloc-
ity, can be adjusted to allocate a greater frequency
sampling for high-velocity reflectors in arterial vessels
moving toward or away from the transducer, as there is
much lower velocity of blood cells in the venous ves-
sels. For instance, instead of allocating the sampling
equally in negative and positive directions, as shown in
Fig. 1.24 (right) of +0.4 m/s to −0.4 m/s, the baseline is
adjusted to scale the velocity from +0.6 m/s to −0.2 m/s Fig. 1.25 Color flow imaging provides a 2D analysis of moving
in the case of arterial flow moving toward the trans- blood by sampling coarse areas within an operator-designated
ducer, which allows a greater fraction of the frequency area on the B-mode grayscale active area. Assignments of
color represent direction of blood velocity. In the image, both
sampling to be assigned to positive frequency shifts arterial and venous flow are present. A Doppler gate for evalu-
resulting from higher blood velocity moving toward ating Doppler spectra is placed for further evaluation of velocity
the transducer. and flow.
Descargado para Anonymous User (n/a) en National Autonomous University of Mexico de ClinicalKey.es por Elsevier en junio 11, 2020.
Para uso personal exclusivamente. No se permiten otros usos sin autorización. Copyright ©2020. Elsevier Inc. Todos los derechos reservados.
CHAPTER 1 Physics of Ultrasound 23
superimposed on the real-time grayscale image. Vas- motion. Estimates of velocity can be limited, and veloc-
culature representing arterial and venous flow is usu- ity variations are not as well resolved as with pulsed
ally within the active area, and color encoding, with Doppler. Aliasing artifacts affect the color flow image
shades of red for blood moving toward the transducer due to insufficient sampling, so areas of high velocity
and blue for blood moving away from the transducer, are represented as reverse flow with a different color
is typically applied. Color flow systems do not calculate assignment.
the full Doppler shift but estimate the shift based on
the similarity of one scan line measurement to another Power Doppler
using an autocorrelation processor to compare the Power Doppler mode is a signal processing method
entire echo pulse of one A-line with that of a previ- used with color flow acquisition that gives up direc-
ous echo. The output correlation varies proportionately tion and quantitative flow estimates but dramatically
with the phase change, which in turn varies propor- increases the sensitivity to any motion by relying on
tionately with the velocity at the point along the echo the total strength and amplitude of all Doppler sig-
trace. Four to eight traces are used to determine the nals, regardless of the frequency shift. A color scale
presence of motion. Direction is preserved through presented with power Doppler reflects the magnitude
phase detection of the echoes. For visualization, the (but not direction) of motion, and by eliminating
grayscale B-mode image and the color flow informa- directionality, the greater sensitivity allows detection
tion must be interleaved. A trade-off of color flow and interpretation of very subtle and slow blood flow.
active image area and frame rate must be considered, Acquisition frame rates are typically slower than with
as a larger area requires more computation time, which color flow acquisition, and “flash artifacts” are com-
leads to slower updates. The color flow image depicts mon with power Doppler imaging mode, arising from
flow direction and vessel boundaries and provides a moving tissues, patient motion, or transducer motion
nice locator for placing a Doppler gate for more precise in conjunction with very high sensitivity. Aliasing is
evaluation of blood velocity, as shown in Fig. 1.25. not a problem, as only the strength of the frequen-
There are several limitations with color flow imag- cy-shifted signals are analyzed, and not the phase. The
ing. Noise and clutter of slowly moving, solid structures name power Doppler can be mistaken for higher depo-
can overwhelm smaller echoes returning from moving sition of energy to the patient, but this is not the case,
blood cells in color flow images. The spatial resolution as the technique involves computational processing.
of the color display is much coarser than the grayscale Images acquired with color flow and power Doppler
image because regions are used to identify areas of are illustrated in Fig. 1.26.
Fig. 1.26 (Left) Abdominal color flow image demonstrating vasculature and directionality of blood velocity;
note color flow scale. (Right) Power Doppler acquisition of the same region, demonstrating increased sensi-
tivity but no directionality of motion.
Descargado para Anonymous User (n/a) en National Autonomous University of Mexico de ClinicalKey.es por Elsevier en junio 11, 2020.
Para uso personal exclusivamente. No se permiten otros usos sin autorización. Copyright ©2020. Elsevier Inc. Todos los derechos reservados.
24 SECTION 1 Principle of Ultrasound
Descargado para Anonymous User (n/a) en National Autonomous University of Mexico de ClinicalKey.es por Elsevier en junio 11, 2020.
Para uso personal exclusivamente. No se permiten otros usos sin autorización. Copyright ©2020. Elsevier Inc. Todos los derechos reservados.
CHAPTER 1 Physics of Ultrasound 25
for the remainder of the PRP the intensity is nearly zero. Cavitation is a consequence of the negative pressures
The temporal peak intensity, ITP, is the highest instanta- (rarefaction of the mechanical wave) that induce bub-
neous intensity in the beam; the temporal average, ITA, is ble formation from the extraction of dissolved gases in
the time-averaged intensity over the PRP; and the pulse the medium. The MI is directly proportional to the peak
average, IPA, is the average intensity of the ultrasound rarefactional (negative) pressure and inversely propor-
pulse. The spatial peak, ISP , is the highest intensity spa- tional to the square root of the ultrasound frequency (in
tially in the beam, and the spatial average, ISA, is the MHz). Cavitation can cause disruption of tissue struc-
average intensity over the beam area, usually taken to be tures by gas bubble implosion and is associated with
the area of the transducer. the formation of free radicals. For ultrasound imaging
Thermal and mechanical indices of ultrasound oper- applications, the MI is typically maintained at low, safe
ation are now the accepted method of identifying power levels, with values substantially less than 1. The United
levels for real-time instruments that provide the opera- States Food and Drug Administration (FDA) has estab-
tor with quantitative estimates of power deposition in lished a maximum MI of 1.9 for diagnostic imaging and
the patient. These indices are selected for their relevance 1.0 for obstetric imaging.
to risks from bioeffects and are constantly updated on
the monitor during real-time scanning. The sonogra- Biological Mechanisms and Effects
pher can use these indices to minimize power deposi- Diagnostic ultrasound has a remarkable safety record.
tion to the patient and fetus consistent with obtaining Despite the lack of evidence that any harm has been the
useful clinical images in the spirit of the ALARA (as low result of ultrasound diagnostic intensities, it is prudent
as reasonably achievable) concept. and indeed an obligation of physicians and sonogra-
phers to consider issues of benefit versus risk when
Thermal Index performing an ultrasound examination and to take all
The thermal index, TI, is the ratio of the acoustic power precautions to ensure maximal benefit with minimal
produced by the transducer to the power required to risk. The American Institute of Ultrasound in Medicine
raise tissue in the beam area by 1°C. This is estimated recommends adherence to the ALARA principles. The
by the ultrasound system using algorithms that consider FDA, which oversees requirements for new ultrasound
the ultrasonic frequency, beam area, and acoustic out- equipment, requires the inclusion of MI and TI output
put power of the transducer. Assumptions are made for indices to give the user feedback regarding power depo-
attenuation and thermal properties of the tissues with sition to the patient. At very high intensities, ultrasound
long, steady exposure times. An indicated TI value of 2 causes consequential bioeffects by thermal and mechan-
signifies a possible 2°C increase in the temperature of the ical mechanisms. The levels and durations for typical
tissues when the transducer is stationary over a given vol- imaging and Doppler studies are substantially below the
ume of tissue. On some scanners, other thermal indices threshold for known undesirable effects. Even though
that might be encountered are TIS (S for soft tissue), TIB ultrasound is considered safe when used properly, pru-
(B for bone), and TIC (C for cranial bone). These quanti- dence dictates that ultrasound exposure be limited to
ties are useful because of the increased heat buildup that only those patients for whom a definite benefit will be
can occur at a bone–soft tissue interface when present in obtained.
the beam, particularly for obstetric scanning of late-term
pregnancies, and with the use of Doppler ultrasound
(where power levels can be substantially higher). TI val- BIBLIOGRAPHY
ues are usually less than 1 for most ultrasound imaging
Abu-Zidan FM, Hefny AF, Corr P. Clinical ultrasound phys-
applications. The limit of TI varies with scanning time; as ics. J Emerg Trauma Shock. 2011;4(4):501–503. https://doi
TI increases, the acceptable scanning time decreases. It is .org/10.4103/0974-2700.86646.
prudent to keep TI values less than 0.7. Anvari A, Forsberg F, Samir AE. A primer on the physical
principles of tissue harmonic imaging. Radiographics.
Mechanical Index 2015;35:1955–1964.
The mechanical index, MI, is a value that estimates Baad M, Liu ZF, Reiser I, Paushter D. Clinical significance of
the likelihood of cavitation by the ultrasound beam. US artifacts. Radiographics. 2017;37:1408–1423.
Descargado para Anonymous User (n/a) en National Autonomous University of Mexico de ClinicalKey.es por Elsevier en junio 11, 2020.
Para uso personal exclusivamente. No se permiten otros usos sin autorización. Copyright ©2020. Elsevier Inc. Todos los derechos reservados.
26 SECTION 1 Principle of Ultrasound
Bakhru RN, Schweickert WD. Intensive care ultrasound: Hangiandreou NJ. AAPM/RSNA physics tutorial for resi-
physics, equipment, and image quality. Ann Am Thorac dents. Topics in US: B-mode US: basic concepts and new
Soc. 2013;10(5):540–548. https://doi.org/10.1513/Annals technology. Radiographics. 2003;23:1019–1033. https://
ATS.201306-191OT. doi.org/10.1148/rg.234035034.
Bushberg JT, Seibert JA, Leidholdt EM, Boone JM. The Essen- Hedrick WR, Hykes DL, Starchman DE. Ultrasound Physics and
tial Physics of Medical Imaging. 3rd ed. Baltimore: Wolters Instrumentation. 4th ed. St. Louis: Elsevier Mosby; 2005.
Kluwer, Lippincott Williams and Wilkens; 2012:500–576. Zagzebski JA. Essentials of Ultrasound Physics. St. Louis:
Chapter 14-Ultrasound. Mosby; 1996.
Descargado para Anonymous User (n/a) en National Autonomous University of Mexico de ClinicalKey.es por Elsevier en junio 11, 2020.
Para uso personal exclusivamente. No se permiten otros usos sin autorización. Copyright ©2020. Elsevier Inc. Todos los derechos reservados.