1✩✩✩✩✩✩✩✩✩✩✩✩✩✩✩✩✩✩✩✩ ✩

1

Sturla H Eik-Nes

Physics and instrumentation

AbstrAct

This chapter provides an overview of the fundamental physical principles that make it possible to produce images of human tissue using sound. The physical laws are explained without the use of complicated formulas. Sound is a mechanical vibration in a medium such as air or human tissue. The upper frequency limit for sound to be heard by humans is 20 kHz. Frequencies above 20 kHz are called ultrasound. Medical images are made with a frequency above 3 MHz. The basic principle for making images of human tissue is to send a pulse into the tissue with a transducer and detect the echoes emerging from structures in the tissue. Imaging may be done in real time by electronic scanning. A variety of sizes and shapes of transducers have been produced for the various applications of ultrasound in medical diagnosis. A proper transducer must be used for a specific task. The ultrasound beam is the essential tool to make images. It must be focused by the user and the image must be properly adjusted with respect to the gain. Measurements can be made and a basic understanding of the resolution in the three planes is necessary for measurements and interpretation of the images. The main artifacts such as edge shadows, attenuation shadows, enhancements and reverberation must be understood. Basic principles of ultrasound scanning must be followed to extract the maximum information from the scan.

Keywords

A-mode, artifacts, B-mode, focus, M-mode, real-time scanning, technical principles of ultrasound in obstetrics and gynaecology, time gain compensation.

IntroductIon

In the practice of clinical ultrasound in obstetrics and gynaecology, it is essential that the examiner has a basic understanding of the physics that makes it possible

Ult

raso

un

d in

ob

stet

rics

an

d g

ynae

colo

gy

2

✩ ✩✩✩✩✩✩✩✩✩✩✩

to produce images of human tissue using sound. In addition, the examiner must be able to handle artifacts properly, know about the basic performance of the instrument and be aware of artifacts, safety and risk factors.

This chapter provides an overview of the fundamental physical principles without the use of complicated formulas to explain the physical laws. The focus is to give the reader an overall understanding of how an ultrasound machine works and the skill to operate the machine and to manage the necessary adjust-ments in order to produce images of high quality for diagnostic use. For in-depth knowledge of the physics of ultrasound, the reader is referred to excellent textbooks. (See selected list at the end of this chapter.)

sound

Sound is mechanical vibrations travelling in a physical medium such as air, water, metal or even human tissue. Whether the airborne vibrations come directly from the source or are reflected, they produce impressions on the eardrums of our ves-tibular organs. We interpret these vibrations as sound.

Sound may be categorized according to various frequency levels:

infrasound (0–20 Hz)•audible sound (20–20 kHz)•ultrasound (>20 kHz)•diagnostic ultrasound (1–20 MHz).•

Humans do not hear the infrasound but other species such as whales, dolphins, elephants, hippopotamuses and rhinoceros do; they use infrasound to communi-cate with other members of their species over long distances. The upper frequency limit for humans is 20 kHz. Frequencies above 20 kHz are called ultrasound. Some species may hear sound frequencies which for humans are categorized as ultra-sound, for example mice (10–70 kHz), dogs (40–60 kHz) and bats (20–200 kHz). There is even some evidence that bats utilize the change in pitch of the echo to determine the relative movement of the object that reflects sound – the Doppler effect. Marine mammals may produce very complex signals ranging from low frequencies for long-range use to high frequencies for local chatting!

short hIstory of the develoPment of ultrAsound In medIcIne

In 1912, the passenger ship Titanic hit an iceberg on its maiden trip crossing the Atlantic from Southampton to New York. In the time that followed, physicists took an interest in using sound to detect large objects submerged in water. Initially their research for that purpose was unsuccessful. During World War I, the French physicist Paul Langevin was responsible for developing the hydrophones needed to detect submarines; this underwater sonar technology resulted in the first sink-ing of a German submarine in 1916. In 1917, Langevin invented the quartz sand-wich transducer which served as the basis for the modern ultrasonic era. Between

Physics an

d in

strum

entatio

n

3

✩✩✩✩✩✩✩✩✩✩✩ ✩

World War I and World War II, the development of sonar (Sound Navigation and Ranging System) and radar (Radio Detection and Ranging) took place. The latter technique used electromagnetic waves rather than ultrasound.

The next important step was the use of ultrasound to detect flaws in metal using high-frequency ultrasound. The metal flaw detectors became increasingly important as World War II was approaching, but were reported after the war.2,4 After World War II, Howry and Bliss, in Denver, started to experiment with sonar equipment and amplifiers from the navy.7 They developed a pulse-echo technique in 1948–49, and later produced cross-sectional images of a human partly submerged in water. At the same time, Wild in Minneapolis developed a breast scanner and actually made a diagnosis of breast lesions with his device.12 The Swedish physician Inge Edler and physicist Helmut Hertz, at the University of Lund, borrowed a metal flaw detector from Kockum's Shipyard in Malmö, Sweden. In 1953, they managed to trace the movements of the human car-diac valves by means of the sound waves emitted and received by their modi-fied instrument.5 This was the start of a new era in cardiology relying on sound technology.6

The next breakthrough was by the Scottish physician Ian Donald, in Glasgow, who conducted the basic research for the development of a machine for clinical use employing ultrasound to make two-dimensional images of human tissue. Donald had served in the Air Force during World War II and his past experience influenced his prototype machine, which consisted of two metal flaw detectors. His Lancet paper of 1958, ‘Investigation of abdominal masses by pulsed ultra-sound’, is considered to be one of the most important for the development of clinical ultrasound.3

Since the late 1950s, the development of ultrasound in medicine in general and in the field of obstetrics and gynaecology in particular has continued in an exponential way. Breakthrough advances have been repeatedly made in spite of claims that the development of ultrasound in medicine has reached its physical limits.

sound, wAves And ProPAgAtIon

Sound is a mechanical vibration in a medium. The medium may be, for example, air, water or human soft tissue. The sound wave propagates through the medium as a longitudinal compression wave. When we think of waves we may picture a stone being thrown into a quiet lake and observe the concentric rings that propa-gate from the centre, or we may think of the waves in the ocean as seen from the shore or from a boat. These waves are transversal waves. Sound waves, however, are longitudinal waves and the medium that they travel through is subject to cyclic variations in pressure as the medium is being compressed or rarefied (Fig. 1.1).

Make a small experiment by putting your index finger on the top of your larynx, then make the sound of a z-z-z. With your finger you will feel the vibrations caused by your vocal cords that are your own sound system, that cause the z-z-z to be heard in the room. You have now produced longitudinal sound waves that travel

Ult

raso

un

d in

ob

stet

rics

an

d g

ynae

colo

gy

4

✩ ✩✩✩✩✩✩✩✩✩✩✩

through the room and cause compression and rarefaction of the air in their path. When the sound waves hit the eardrums of someone in the room, the process is reversed and causes the eardrums to vibrate and the person will hear your z-z-z.

The sound wave is a longitudinal wave caused by compression and rarefaction of a physical medium in the direction of the movement of the wave.

This sound wave may further be described by intensity and frequency.If you have a piano, you can carry out a small experiment in your living room

by hitting A above middle C. You will hear a chamber tone with a frequency of 440 Hz. If you move up one octave on your piano and hit A, you will hear it at a frequency of 880 Hz. If you move up one more octave to the next A, you will hear an A note with the frequency of 1760 Hz.

The frequency tells us about the degree of highness or lowness of a tone. The fre-quency is the number of vibrations per second that produce the sound.

Hit the A on your piano very lightly and you will barely hear the chamber tone of 440 Hz; hit the key with force and you will hear the same chamber tone with the frequency of 440 Hz, but much louder. This tells us that the same tone may differ in intensity or loudness.

The intensity tells us something about the loudness or strength of the sound signal.A sound wave travelling in a medium produces compression and rarefaction of

the medium as shown in Figure 1.1. The velocity of propagation of the sound wave is dependent on the medium and is 330 m/s in air, 1480 m/s in water, 1589 m/s in

Compressionλ

Pres

sure

Decompression

Distance

Distance

Moves with wave velocity, c

fig. 1.1 (Upper panel) A schematic illustration of a sound wave as it travels in a medium causing periodic compressions and rarefaction of the medium. (Lower panel) The dislocation of the particles.

Physics an

d in

strum

entatio

n

5

✩✩✩✩✩✩✩✩✩✩✩ ✩

muscle and 3500 m/s in bone. The hardness or stiffness of the medium is the main factor determining the propagation velocity of sound.

Ultrasound machines are now standardized and calibrated to use 1540 m/s as the speed of sound in human tissue. Based on the propagation of the sound wave in a particular medium (v) with a particular frequency (f), we arrive at the first important equation for the wavelength λ:

(1)

A chamber tone (440 Hz) has a wavelength of 0.75 m, propagating in air at the velocity of 330 m/s. It is obvious from equation 1 that the wavelength will vary with the frequency and velocity of sound in the tissue. The higher the frequency, the shorter the wavelength; the higher the velocity of sound, the longer the wavelength. Because the speed of sound in human tissue has been standardized at 1540 m/s in the equation, the wavelength will vary with the frequency (Table 1.1).

The higher the frequency of ultrasound in human tissue, the shorter the wavelength.An ultrasound wave with a frequency of 5 MHz (M is the Greek abbreviation

for mega which means big, but used in acoustics it means million) has a wave-length of 0.31 mm.

It is important to understand what really happens when a sound wave moves through the medium. A scene we all are familiar with will demonstrate the prin-ciple (Fig. 1.2).

When a sound wave propagates through a medium, the wave moves while the medium remains in place. Thus, when ultrasound propagates through human tissue, it is the wave that moves, not the tissue.

Let's go back to the sound waves. Low-frequency sound (a human voice, music) will spread all over a room. You can easily hear the voice of a person talking with his back turned to you. Very high-frequency sound behaves like light – it moves like a beam along a straight line.

High-frequency ultrasound propagates through tissue in a relatively narrow beam and may be focused by acoustic lenses.

In order to make a simple ultrasound machine, we need to be able to produce high-frequency sound. In the 1880s the Curie brothers discovered the piezoelec-tric effect which implies that a crystal, for example a quartz crystal, will produce an electrical current if subject to mechanical pressure. Conversely, an electrical cur-rent that is applied to a quartz crystal will cause the crystal to change its shape. The change in shape will have an impact on the surrounding medium. If alternating

λ =v

f

frequency (mhz) wavelength (mm)

3.5 0.44 5 0.31 8 0.1910 0.15

table 1.1 Various ultrasound frequencies and the corresponding wavelength

Ult

raso

un

d in

ob

stet

rics

an

d g

ynae

colo

gy

6

✩ ✩✩✩✩✩✩✩✩✩✩✩

current is applied to the crystal, the crystal will repeatedly change its shape and the movements of the crystal will produce a wave transmitted through the medium.

By using a piezoelectric material (quartz crystal) it is possible to produce high- frequency sound waves that emerge from the crystal into human tissue. The same crystal can be made to pick up the echoes emerging from the depth of the tissue. Such echoes will have an impact on the crystal that produces an electric pulse that we may detect and process further.

If you have been at an outdoor rock concert in front of a full-blast subwoofer, you will have experienced the impact that sound can have on your body, in par-ticular on your air-filled chest cavity. Imagine the sound level scaled down to an impact you cannot feel and then a very sensitive instrument introduced to detect the sound waves; then you have a demonstration of the basic principle of receiv-ing low-impact echoes. Making images with sound is about sending and receiving sound waves in the form of a pulse (Fig. 1.3).

We now have enough knowledge to make a one-dimensional ultrasound image of the fetal skull the way it was done in the late 1950s and early 1960s. It was called A-mode (A stands for amplitude) (Fig. 1.4).

In the early days of the clinical use of ultrasound, A-mode technology made it possible to measure the fetal biparietal diameter and the conjugata vera, to locate the placenta, including placenta praevia, and to diagnose polyhydramnion, detect the fetal heart activity, diagnose a molar pregnancy and a variety of other diagno-ses. The interpretation of such images was difficult and required extensive train-ing and imagination of the examiner. Still, sophisticated diagnoses were made by dedicated pioneers.9

Wave motion

fig. 1.2 The person on the shore throws a stone into the water. The stone creates waves in the form of concentric rings that approach the cork. Instead of being ‘pushed away’ the cork moves up and down as the wave passes by.

Physics an

d in

strum

entatio

n

7

✩✩✩✩✩✩✩✩✩✩✩ ✩

The natural step forward was to make two-dimensional images. The strength of the echoes was then displayed as a white dot instead of as an amplitude; the higher the intensity of the returned echo, the larger the dot. This was called B-mode (B stands for brightness). In a one-dimensional system, these signals were impos-sible to interpret but moving the transducer in a plane across the area to be exam-ined (scanning) during sending and receiving made it possible to display all the echoes emerging from structures in that plane. Together, these were converted into a relatively easy-to-read two-dimensional image (Fig. 1.5). This manual scanning made it much easier to produce and interpret two-dimensional images produced with ultrasound. The image quality was further improved by the devel-opment of the analogue scan converter, so that grey scaling could be applied as well as scaling of the image and calliper movements on the screen.

The next technical step was to produce real-time two-dimensional images. This was achieved mechanically in the 1960s by Krause and Soldner in Erlangen, Germany.10 A more sophisticated way was to align a set of crystals to make a linear transducer, described by Nicolaas Bom in Rotterdam, in 1971.1 The prin-ciple was further developed by Martin Wilcox who produced a clinically most successful real-time scanner in 1972 (Fig. 1.6).

The principle of displaying the returned signals appropriately is simple: the speed of sound is known and the time from when a pulse is emitted until it comes back can be calculated. It is obvious that each submitted pulse will hit many structures in the path of the beam, thus many echoes will be returned separated by a short time interval.

Electronic real-time scanning implies that the transducer sends a pulse, and then it switches to the listening mode. A linear transducer may typically have 196 or more crystals aligned in a row. Typically crystals number 1–50 are fired, and then number 2–52, etc. The examiner is presented with an image frame rate of approximately 30 per second, which for the human eye will make the on-screen image appear flicker free with movements in real time.

fig. 1.3 (Bottom) An electric current is applied to the transducer and a pulse is sent out. (Top) A pulse is received and generates an electric current that can be displayed by the instrument. The stronger the returned pulse (echo), the higher the amplitude of the electric current.

Ult

raso

un

d in

ob

stet

rics

an

d g

ynae

colo

gy

8

✩ ✩✩✩✩✩✩✩✩✩✩✩

fig. 1.4 A-mode. A single ultrasound beam is sent through the fetal skull and, in sequence reflected from the parietal bone closest to the transducer, the falx cerebri, the skull bone distal to the transducer and, finally, the posterior uterine wall. Depending on the strength of the returned pulses (echoes), the quartz crystal will generate a high- or low-amplitude current.

Physics an

d in

strum

entatio

n

9

✩✩✩✩✩✩✩✩✩✩✩ ✩

Finally, we need to understand the physical principle of M-mode (M stands for motion). M-mode is used to trace the movement of a structure. For example, tracing the movement of a heart valve or the movement of the atrial wall and the ventricular wall of the fetal heart simultaneously on the very same image makes it possible to discriminate a dissociation of the rhythm, i.e. supraventric-ular tachycardia, various kinds of AV block, etc. During an M-mode recording, we register the movements of the echoes along one single line in our image (the y-axis) while time runs along the x-axis. The principle is easy to understand if we imagine that we put a long paper strip on our desk, hold a pen against the paper and move the pen up and down while pulling the paper strip in a direction perpendicular to the movement of the pen. In our example, the pen represents the moving echoes and the up-and-down movement of the pen will result in a curved line on the paper reflecting the movements of the pen. An M-mode scan is shown in Figure 1.7.

one trAnsducer for eAch PurPose

A variety of sizes and shapes of transducers have been produced for the various applications of ultrasound in medical diagnosis. Transducers have various sizes of ‘footprints’, i.e. the part of the transducer that touches the skin or other tissue. Transducers with a small footprint are necessary in, for example, cardiology, for sending a beam between the ribs to reach the heart as a target organ. To reach the heart and thoracic aorta, even an oesophageal transducer may be used; in urology and proctology, the prostate or lower part of the intestines is reached by inserting a transducer into the rectum. The gastroenterologist may examine the liver from the surface of the abdomen or insert a slim transducer through the gastric scope to reach the surrounding organs including the ductus pancreaticus and the pancreas.

fig. 1.5 Twin pregnancy. B-mode image obtained in 1964 by Diasonograph, Nuclear Enterprises Ltd, Edinburgh, UK. The image is made by ‘compound scanning’, i.e. by rocking the probe back and forth during the process of moving the scanning arm slowly across the pregnant abdomen. Reproduced by permission from Bertil Sundén.11

Ult

raso

un

d in

ob

stet

rics

an

d g

ynae

colo

gy

10

✩ ✩✩✩✩✩✩✩✩✩✩✩

In vascular surgery, imaging through a catheter has been developed for target organs such as the neck vessels and coronary arteries. In the field of obstetrics and gynaecol-ogy, curvilinear transducers are extensively used for transabdominal examination (Fig. 1.8). The shape of the transducer fits well to the pregnant and non-pregnant abdomen, the footprint is small, while the view deep in the tissue is wide due to the sector-shaped image. The use of a convex transducer also reduces the effect of reverberations and wave front aberrations (see later). Transducers designed for transvaginal scanning make the early pregnancy and the non-pregnant uterus acces-sible at a close range; thus, they are widely used in gynaecology and obstetrics.

Transducer technology has become complex. The essential unit, the sound-emit-ting crystal, was made of natural materials such as quartz. Nowadays most of the crystals are made of artificial ceramics mixed with plastic materials with various

fig. 1.6 (Upper panel) The ADR linear scanner (in Europe manufactured under the name of ADR-Kranzbühler, image by courtesy of the company, 1980). (Lower panel) The basic principle of scanning in real time. The crystals fire the sound beams, which travel into the human tissue, hit structures and are reflected. The reflected echoes are picked up and displayed accordingly on a screen.

Physics an

d in

strum

entatio

n

11

✩✩✩✩✩✩✩✩✩✩✩ ✩

forms of damping material to produce a clean pulse and a pulse of short duration. The electrical excitement is made through thin silver electrodes connected to the ceramic material. The basic principle for producing a pulse wave and receiving an echo, which generates a current, remains the same, as illustrated in Figure 1.8.

the ultrAsound beAm

NEAR FIELD AND FAR FIELD

Ideally, an ultrasound beam would emerge from a crystal, be narrow and circu-lar and shoot into the tissue along a straight line. Then it would return along the same line from structures it may hit, to the very same crystal, which would be excited by the echoes and produce an electrical current. In real life, the beam is not ‘narrow and circular’ but advanced engineering has, over time, worked to

A-mode M-mode

Movement with time

Movement with time

TimeGrey-scale amplitudealong beam at fixed time

Dep

th

fig. 1.7 The principle of M-mode. The basic principle is described in the text.

fig. 1.8 Sector, curvilinear, linear and transvaginal transducers.

Ult

raso

un

d in

ob

stet

rics

an

d g

ynae

colo

gy

12

✩ ✩✩✩✩✩✩✩✩✩✩✩

modify the beam towards the ideal form. Put simply, the beam has a near field and a far field. In the near field we may influence the shape of the beam by focus-ing. In the far field we cannot do that. When we make our images we are operat-ing in the near field. So from an imaging point of view, we would like the beam to have a long near field (Fig. 1.9).

The depth (d) at which the transition of the beam from the near field to the far field takes place is given by equation 2. r is the diameter of the circular transducer:

(2)

This equation tells us that the near field is relatively long if the diameter of the circular transducer is large and/or the wavelength (λ) is short, i.e. the frequency is high. It follows that the near field is relatively short if the transducer has a small diameter and/or the wavelength is long, i.e. the frequency is low.

FocUSINg

This brings us to the next important feature, which is the focusing of the beam. The required effect of focusing the beam is to reduce the width of the beam. Focusing may be achieved by employing lenses in various forms.

(3)

Figure 1.10 shows the trade-off of having a narrow beam width as an effect of focusing: an increased divergence of the beam distal to the focal distance.

Considering equations 1–3, we may conclude that a focused transducer with a large diameter (aperture) and a high frequency (short wavelength) will provide a narrow beam in our region of interest (at the focal distance). So why do we not settle for transducers with a large aperture and a high frequency?

The quick answer is that a large aperture may not be acceptable for a particular application and high-frequency ultrasound is absorbed to a greater extent than low-frequency ultrasound. The range of a relatively low-frequency transducer is longer than for a relatively high-frequency transducer.

d = r 2

λ

BW = ×F

r

λ2

Near field Far field

Transducer

d

2r

fig. 1.9 Schematic illustration of an ultrasound beam emerging from a transducer with a circular surface with a diameter 2r. The beam has a near field reaching into the depth of d and a far field.

Physics an

d in

strum

entatio

n

13

✩✩✩✩✩✩✩✩✩✩✩ ✩

The solution is to use high frequency if we are looking at structures close to the transducer and low frequency if we are looking at structures further away.

Let us go back to the ultrasound beam. Ideally, one would like the ultrasound beam to be thin and round, and shoot into the tissue along a straight line, hitting structures which cause echoes that return to the transducer along the same line. Then only structures in the thin path of the beam would cause echoes. We have learned that this is not so. The beam has a near field where we may manipulate the beam and a far field where the beam diverges due to diffraction, where it is not possible to manipulate the beam. The beam has a main lobe and side lobes (Fig. 1.11). The side lobes may be considered ‘skirts’ around the main lobe body. When such a complex beam is shot into the tissue, all the structures hit by the

Near field Far field

Transducer Lens BW

F

2r

fig. 1.10 Schematic illustration of an ultrasound beam emitted from a transducer with a circular surface with a diameter 2r. The focal distance is set at F and the beam width (BW) is the effect of the focusing.

Side lobeMain lobe

fig. 1.11 Schematic sketch of an ultrasound beam, demonstrating the main lobe and the side lobes.

Ult

raso

un

d in

ob

stet

rics

an

d g

ynae

colo

gy

14

✩ ✩✩✩✩✩✩✩✩✩✩✩

main lobe and structures which in reality are located on the side of the main lobe, but within the side lobes, will cause echoes to be returned to the transducer and be displayed along the centre of the imaginary line through the centre of the main lobe. This will cause a ‘smear-out’ effect of the image.

The presence of side lobes in addition to the main lobe reduces the quality of our image. Structures outside the main lobe will be picked up by the side lobes and on the final image they will be displayed along the centre line through the main lobe.

Improving the overall beam quality is accomplished through the focusing pro-cess which may be achieved in various complex ways. One technique, dynamic focusing, may help us understand the principle of focusing. One submitted pulse may cause many returned echoes which hit the transducer surface over a time period, depending on how far the echoes have travelled on their way down to the various reflectors and then back. Focusing is a process that may be done on the way out and on the return of echoes. Focusing on the returned echoes is always done. Since we know when a pulse has been transmitted, we may focus the returned echoes by changing the focus level in the tissue at certain time intervals following the transmission of the pulse. This will cause echoes, which originate from a depth of, for example, 2, 4, 6, 8 and 10 cm, to be focused separately on return. Thus, the focusing process will affect the area between 2 and 10 cm, in the example above.

Additionally, we may focus our area of interest especially on the way out to obtain the highest image quality possible in the specific area where we are look-ing. Arrows along the side of the image indicate the manually set foci. Optimum quality is usually achieved employing two to three foci in the area of interest.

The process of directing the focus of our beam to the area we are looking is one of the most important manual adjustments we make during ultrasound scanning. Unfortunately, focusing is one of those manual adjustments which are most often forgotten, a practice that exemplifies a lack of technical understanding of the person performing the scanning.

resolutIon

To be able to interpret our image, define discrete structures and make precise measurements on an ultrasound image, we need to understand the basic prin-ciples of resolution.

Resolution is defined as the smallest distance we can have between two structures and still be able to distinguish them as two separate structures.

On a two-dimensional ultrasound image, we have an axial plane, a lateral plane and an elevation plane (Fig. 1.12). The resolution in these three different planes is determined by various physical laws that we have to understand to optimize the adjustment of our machine settings, select the best transducer for our purpose and make measurements as precise as possible.

The axial resolution may be called the range resolution or the radial resolution. The resolution in the axial plane is the best of the three. The axial resolution is mainly determined by the length of the transmitted pulse. A ‘pulse’ always con-sists of a few oscillations in spite of effective damping factors. The absolute length

Physics an

d in

strum

entatio

n

15

✩✩✩✩✩✩✩✩✩✩✩ ✩

of a pulse may therefore be reduced by increasing the ultrasound frequency. The principle of the axial resolution is demonstrated in Figure 1.13. The total pulse length of a 5 MHz pulse is typically shorter than that of the 2.5 MHz pulse.

A good axial resolution requires a short pulse. Several factors may contribute to a short pulse: one of them is the wavelength. A high frequency (i.e. short wavelength) will make the pulse relatively short and improve the axial resolution.

Transducer

Axial planeAzimuth plane

Elevationplane

fig. 1.12 The three planes on an ultrasound image: the axial, the azimuth and the elevation plane.

5 MHz

2.5 MHz

5 MHz

2.5 MHz

5 MHz

2.5 MHz

fig. 1.13 In the upper part, a 5 MHz pulse is shown travelling towards a target, which may be a blood vessel. The pulse is short enough to be able to hit the anterior and posterior walls separately, thus two separate echoes will be reflected and make two separate dots on the screen when they hit the transducer. Below, the 2.5 MHz pulse is longer and the echoes from the anterior and posterior walls of the vessel will overlap, so only one large dot will be displayed on our screen. The 2.5 MHz pulse was not able to resolve the two vessel walls as two separate structures.

Ult

raso

un

d in

ob

stet

rics

an

d g

ynae

colo

gy

16

✩ ✩✩✩✩✩✩✩✩✩✩✩

The lateral resolution affects measurements across the azimuth plane, which is perpendicular to the direction of the beam. The lateral resolution is governed by different physical laws from the axial resolution and is poorer than the axial reso-lution. Among the factors that affect the lateral resolution are the quality of the beam and the size of the side lobes (see Fig. 1.11). In the process of optimizing the beam quality, the aim is to have a thin main lobe and small side lobes.

The lateral resolution perpendicular to the direction of the beam is poorer than the axial resolution. Measurements made in the axial direction are more precise than those made across the image perpendicular to the beam.

meAsurement

Generally, when we measure a distance in the axial direction, we put one electronic calliper on an echo and move the next calliper to another echo to assess the distance between the two. However, we are not actually measuring the distance between the two, but rather the time it takes for a pulse to travel from the transducer to the struc-ture closest to the transducer and to the structure further away. So when we measure a distance, our calculations are based on time rather than on a physical distance.

We have to take into account that of the two, axial resolution is better than lat-eral resolution. If we measure in the plane perpendicular to the beam, the beam quality will influence our measurement. A relatively thick beam will make the endpoint of a structure appear blurred and make the distance between two points appear slightly larger than in reality. This phenomenon has a consequence for the measurements across the screen, for example the occipitofrontal diameter of the skull and even the femur length.8

tIme gAIn comPensAtIon

When a pulse propagates through the tissue, it will gradually lose its energy. This loss is caused mainly by power absorption and to a smaller extent by reflection, scat-tering and geometric spread. This process takes place as the pulse travels away from the transducer and as the echo is on its way back to the transducer. The absorption of ultrasound energy increases with increasing frequency. The attenuation causes the reflected echoes from structures deep in the tissue to be weaker than those emerging from nearby structures. If we do not compensate for this phenomenon, our image will appear imbalanced (Fig. 1.14). The speed of sound in the human tis-sue is constant; the echoes emerging from the deeper areas arrive later than those from the upper structures. Thus, we may compensate for the loss of power from the late-arriving echoes by inserting a time variable gain in the receiver amplifier. This is called time gain compensation (TGC). The basic TGC is preset in modern machines, but we may have to adjust manually to fine-tune our image. Usually it is possible to make an overall adjustment of the gain as well as adjustments affect-ing the local area ranging from the near to the far field of the image. The setting of the TGC also affects our measurements and it is an important part of the training to learn how to set it correctly. A TGC adjusted too high will produce blurry edges and measurement of distance between structures will be longer than in real life.

Physics an

d in

strum

entatio

n

17

✩✩✩✩✩✩✩✩✩✩✩ ✩

The fine-tuning of our image using the TGC is one of the most important adjustments we make. The grey-scale level of the image ought to appear well bal-anced from the upper to the lower part of the image. The adjustment must aim at achieving the full register of grey tones between the black areas and the white highlights. The setting of the TGC has an influence on our measurements.

ArtIfActs

Artifacts in ultrasound imaging may be distortions or any form of incorrect appear-ance affecting an image and giving misleading information as we try to interpret from the image. The imaging process using ultrasound technology may cause numerous artifacts that we have to be aware of. Some of the main artifacts are as follows:

Edge shadows•Attenuation shadows•Enhancement•Reverberations.•

EDgE SHADoWS

In obstetrics, edge shadows are mainly observed during scanning of the fetal head. When the sound enters a round structure such as the fetal head it emerges from tissue with a velocity of 1540 m/s through the bone of the fetal skull that has a sound velocity of ≈3000 m/s. The sound will then be refracted and leave a shadow-like impression on both sides of the fetal skull (Fig. 1.15).

ATTENUATIoN SHADoWS

Bone absorbs ultrasound and the echo amplitude will then be reduced behind ossified structures. This is frequently observed during fetal heart scanning when the image of the heart may be in the shadow of the ribs or the vertebrae. In gynaecology, dense structures such as myomas may to a lesser degree reduce the

fig. 1.14 A section through the planum biparietale. The area close to the transducer is correctly adjusted while the distal area has hardly visible low-energy echoes as a consequence of the insufficient compensation for the attenuation of sound emerging from the deeper sections of the tissue.

Ult

raso

un

d in

ob

stet

rics

an

d g

ynae

colo

gy

18

✩ ✩✩✩✩✩✩✩✩✩✩✩

amplitude of the sound. Such shadows may give us information about the structure that is causing the shadow.

ENHANcEMENT

Enhancement is the opposite of attenuation shadow. The phenomenon may be seen behind cysts (Fig. 1.16). This artifact may also be used to characterize the structure causing the enhanced area.

fig. 1.16 Simple cyst demonstrating the enhancement artifact. The sound that is passing through the cyst is not attenuated in the same degree as the sound passing through the tissue on the right and left side of the cyst. Therefore, the amplitude of the sound immediately below the cyst is higher than on the sides and consequently it looks as if the area below the cyst has been enhanced by selectively turning up the time gain compensation.

fig. 1.15 Edge shadows. on both sides of the fetal skull, the ultrasound beam is refracted and then leaves a shadow below.

Physics an

d in

strum

entatio

n

19

✩✩✩✩✩✩✩✩✩✩✩ ✩

REVERBERATIoNS

The artifact referred to as reverberation or multiple reflections is common and may distort the image in several ways. The basic principle of making an image using sound is to send a pulse, wait for the pulse to return as an echo and then a dot is put on the screen corresponding to the time the pulse has taken to travel on its way down to the reflecting structure and back again. A pulse may also be reflected back and forth between interfaces before returning to the trans-ducer. The extra travel time this process takes will cause the false echoes to arrive later than echoes emerging directly from the original structure so that then sev-eral lines on the image may present themselves as copies of the original (Fig. 1.17). Such reverberations may easily be recognized. Layers of fat may also cause reflections and reverberations in the image that presents itself as a diffuse cloud of noise and is thus not so easy to recognize as artifacts. The lower mechanical impedance of sound in fat (sound velocity ≈1420 m/s) and muscle tissue (sound velocity ≈1560 m/s) may cause reverberations.

Reverberations may be complex in their appearance and not always easy to detect. Using a curved array transducer may reduce the effect of reverberations. The echoes are scattered out of the field, which causes the curved array to have a good near-field view.

fig. 1.17 Two examples of reverberations. In the upper panel (A) the main echo schematically is represented by a blood vessel. A ‘copy’ of these echoes may be found as reverberations at a lower level. Fatty tissue may also cause reverberations which may show up as a diffuse haze (B).

Object Image

Main echo

Reverberation

Reverberation

Main echo

A

B

ImageObject

Ult

raso

un

d in

ob

stet

rics

an

d g

ynae

colo

gy

20

✩ ✩✩✩✩✩✩✩✩✩✩✩

References

1. Bom N, Lancée CT, v Zwieten G, Kloster FE, Roland J. Multiscan echocardiography I. Technical description. Circulation 1973;48(5):1066–1074

2. Desch CH, Sproule DO, Dawson WJ. The detection of crack in steel by means of supersonic waves. J Iron Steel Inst 1946; 153:319

3. Donald I, Wicar WA, Brown TG. Investigation of abdominal masses by pulsed ultrasound. Lancet 1958;1:1188

4. Firestone FA.The supersonic reflectoscope, an instrument for inspecting the interior of the solid parts by means of sound waves. J Acoustic Soc America 1946;17:314

5. Edler H, Hertz CH. The use of ultrasonic reflectoscope for the continuous recording of movements of heart walls. Kgl Fysiograph Saellskap Lund Förh 1954;40:23

6. Edler I. Ultrasound cardiography. The diagnostic use of ultrasound in heart disease. Acta Med Scand 1955;308(suppl):32

7. Howry, DH, Bliss WR. Ultrasonic visualisation of soft tissue structures of the body. J Lab Clin Med 1952;40:579

8. Jago JR, Whittingham TA, Heslop R. The influence of ultrasound scanner beam width on femur length measurements. Ultrasound Med Biol 1994;20(8):699–703

9. Kratochwil A. Ultraschalldiagnostik in Geburtshilfe und Gynäkologie. Georg Thieme Verlag, Stuttgart, 1968

10. Krause W, Soldner R. Ultraschallbildverfahren (B-Scan) mit hoher Bildfrequenz für medizinische Diagnostik. Elektromedica 1967;4:1

11. Sundén B. On the diagnostic value of ultrasound in obstetrics and gynæcology. Acta Obstet Gynaecol Scand 6(suppl):114

12. Wild JJ, Reid JM. Application of echo-ranging techniques to the determination of structure of biological tissues. Science 1952;28:226–230

Further reading

Angelsen B. Ultrasound Imaging. Waves, Signals, and Signal Processing. Basic Principles, Wave Generation, Propagation, and Beam forming in Homogenous Tissue. Vol I. Emantec, Trondheim, Norway, 2000. www.ultrasoundbook.com

Angelsen B. Ultrasound Imaging. Waves, Signals, and Signal Processing. Propagation and Scattering in Homogenous, Nonlinear Tissue with Contrast Agent. Imaging and Doppler Measurement. Vol II. Emantec, Trondheim, Norway, 2000. www.ultrasoundbook.com

Hatle L, Angelsen B (eds). Doppler ultrasound in cardiology. Physical principles and clinical applications. Lea and Febiger, Philadelphia, 1986

Kremkau FW. Diagnostic ultrasound. Principles and Instruments, 7th edn. WB Saunders, Philadelphia, 2006

Maulik D (ed). Doppler ultrasound in obstetrics and gynecology. Springer, New York, 1997Woo J A short history of the development of ultrasound in obstetrics and gynecology.

www.ob-ultrasound.net/history1.html