Ultrasound instruments have an impressive array of switches, dials, and buttons that can be daunting to the novice (Fig. 1.2). Even finding the on/off button to the instrument, which is invariably in a covert location to avoid inadvertent deactivation, adds to the challenge. Once the instrument is turned on, however, users quickly find that the instrument is quite intuitive and can be quickly put to use. The first step is to identify the active transducer on the instrument and then to coat it with coupling gel. It can then be placed over the wrist as shown in Figure 1.3. Note that every linear array transducer has a marking on one side, which corresponds to a marking on the display screen. The marking on the display should always be in the upper left corner of the screen, and the probe should be positioned such that the marker on the probe is pointed cephalad to the patient for sagittal imaging and to the ultrasonographer’s left side for cross-sectional imaging. This will produce images that are oriented in the same fashion as computed tomography (CT) or magnetic resonance imaging (MRI). However, it should be noted that this convention has not been routinely followed in neuromuscular imaging, so not all of the older images in this text are oriented in this fashion. In particular, sagittal imaging of the median nerve at the wrist frequently has been depicted in the opposite fashion in the literature, with the left side of the display screening showing the distal portion of the wrist, but that is an undesirable orientation, based on ultrasound convention, and should be avoided.² See Box 1.1 for other general considerations when performing neuromuscular ultrasound.

Fig. 1.2 Control panels of several different ultrasound instruments, all circa 2010. A, Philips IU22. B, Biosound Esaote MyLab gold. C, GE Logiq. Although modern instruments are digital, instrument controls can either involve continuous adjustment or incremental steps. The time gain compensation panel on each instrument is indicated by arrows.

Fig. 1.3 Probe on wrist, the fingers cover the edge indicator on the transducer, and the transducer is held to obtain a cross-sectional image of the median nerve at the wrist.

Box 1.1 General Considerations for Neuromuscular Ultrasound

1. Use a linear array transducer with a frequency ≥ 12 MHz.

2. Set up the exam room so the ultrasonographer can see bothination the patient and ultrasound device in the same view (see Fig. 1.25).

3. Always position the marker on the transducer, which corresponds to the upper left corner of the ultrasound display screen, to the ultrasonographer’s left side with cross-sectional imaging and to the patient’s head with sagittal imaging.

4. During the study, the depth, focus, gain, and time gain compensation should be adjusted to optimize imaging of the structure of interest.

5. If available, Doppler should be used to confirm flow in vessels.

6. Obtain and save images of each structure of interest in at least two views. For example, in patients with carpal tunnel syndrome, both cross-sectional and sagittal images of the median nerve should be assessed and saved.

7. Label structures of interest using the annotation capabilities of the ultrasound device, and save key images. These images can be printed, saved to the hard drive of the ultrasound device, or uploaded to a picture archiving and communication system.

Fig. 1.25 Image showing the ease of transition from electromyography (EMG) to ultrasound in the care of patients and education of house staff.

When imaging the wrist, the base of the palm should rest comfortably on the patient’s forearm or other supported structure. An image of the contents of the carpal tunnel, and possibly the radial and/or ulnar artery should appear. The width of the image corresponds to the footprint of a linear probe (Fig. 1.4) or the radial angles of the probe if it is curved (see Fig. 4.2 and Fig 4.14). The coupling gel, in addition to enhancing sound penetration, allows the probe to slide left or right or distal or proximal to find the optimal image of the tissue in the neighborhood of the initial contact of the probe. The probe can also, with slight pressure, be angled proximally or distally. In general, structures are viewed with the probe perpendicular to the skin surface, but angular adjustments are of considerable interest as is discussed later. No more pressure than necessary should be used to scan; excessive pressure may deform structures of interest and can be tiring to the ultrasonographer and uncomfortable for the patient (Fig. 1.5). To help others orient to an image, it can help to include a recognizable landmark as well (vascular bifurcation, bone edge, etc.).

Fig. 1.4 This figure shows the difference between probe footprints. The top row shows a hockey stick (small footprint) probe on the wrist, and the cross-sectional image it generates. A sagittal image at the wrist from this probe is also included. The second row shows the standard probe and its cross-sectional image and sagittal image at the wrist for comparison. The bottom image is a direct comparison of the two types of probes. Note the extended field of view of the larger probe, which includes both the radial (R), and ulnar (U) arteries, and associated veins, the median nerve (M), and multiple tendons, whereas the smaller probe can only encompass the radial artery by sacrificing the lateral border of the median nerve. Notice how much more of the median nerve (and underlying bones in the hand) are apparent with the standard than hockey stick probe on the sagittal views. Note the reverberation artifact beneath the radius in the second row. Also note in the sagittal images that distal is to the left, as these are older images that do not follow standard convention.

Fig. 1.5 These images show the effects of compression on nerve and muscle. A, Both images are axial views of the median nerve at the distal wrist crease. Note that with minimal downward pressure (left image) and maximal downward pressure (right image) there is little change in nerve thickness or shape even though the distance between the nerve and radius is less with higher pressure. B, Both images are axial at the middle of the metacarpal (arrows) through the abductor pollicis brevis muscles. The left image is with minimal transducer pressure, the right image is with maximal transducer pressure. Note that the pressure decreases abductor pollicis brevis thickness by almost 50%. Maximal pressure at the carpal tunnel causes less deformation than over the abductor pollicis brevis because tendons and nerves are less compressible/displaceable than skleletal muscle.

Several simple adjustments can be made to change the display once the desired image is obtained. One knob on the ultrasound instrument panel adjusts the depth of the field of view.³ The depth should be sufficient to encompass the structure(s) of interest. The display screen is calibrated to indicate depth and serves as a useful reference in images for publication (see Fig. 1.4). The zoom feature of an instrument allows the user to focus in on the detail of interest, expanding it to encompass the screen with a slight improvement in resolution. Unlike MRI and CT, which are designed to automatically record standard body slices, ultrasound requires the operator to scan using multiple probe positions and angles and operating controls to capture the most informative two-dimensional (2D) images for documentation. Greater depth of field (low magnification) and zoom (higher magnification) are analogous to lens power on a microscope—lower magnification is useful for orientation and screening for abnormalities, and higher magnification is used to capture the essence of the pathologic change. If the pathology is diffuse, lower magnification (greater depth of field, as in the case of diffuse muscle pathology) may be preferred.

The brightness of the image can be controlled by the power and gain dials on the instrument.³ The power dial controls the amount of sound energy transmitted to the tissue and the gain dial the amplification of the sound echoes that return. These dials have similar but not identical functions and need to be set to optimally display the structures of interest. The dials have an analogy with photography: The power dial in this case controls the strength of the flash, and the gain dial in essence controls the film speed (with faster, e.g., more sensitive, film speeds being equivalent to increased gain). In electrodiagnosis, the power dial controls stimulus intensity⁴, and gain is used to display the amplitude of the nerve conduction study response. In both ultrasound and electrodiagnosis, excessive intensities do little to enhance the outcomes of the study.

The focal zone, typically indicated by horizontal marker(s) on the display screen, helps with resolution.³ The level of focal zones (which are horizontal) should encompass the structures of interest (Fig. 1.6). The freeze button on the instrument will stop the real-time action of the machine and display a single frame. Most instruments store (in an adjustable fashion) several seconds of data at any given time, so it is possible, once an image is frozen, to rewind slowly through multiple previous frames to capture the optimal image for storage. Of interest, the rewind and replay features of ultrasound instruments have been an industry standard for many years, yet this simple function has only recently become available on electromyography (EMG) instruments. Any frozen image can be saved for later viewing, and a variety of postprocessing capabilities for analysis and labeling are present on many instruments as well.

Fig. 1.6 Images of the median nerve in the forearm, at the level of the bifurcation of the brachial artery deep to the pronator teres. In the image on the left the focal zone is at the level of the median nerve and on the right, the focal zone is just below the surface (large flanking black arrows). Note that the detail of the structure of the median nerve is better defined on the left (upper green arrow), as is the outline of the brachial artery (lower green arrow) than it is on the right.

The Transducer

The generic term transducer refers to something that converts one type of energy into another. Biologic receptor organs are all transducers, converting different types of physical energy into electrical impulses in nerves; the retina converts light, the cochlea sound, nerve endings heat, and so on. Muscle is a transducer that operates in reverse to a sensory transducer because it converts electrical energy into mechanical energy. It does this, aided by ample reserves of adenosine triphosphate (ATP), with remarkable leverage. The act of lifting heavy objects begins as a few central nervous system (CNS) neuronal discharges on the order of microvolts, and these discharges are quickly transformed into hundreds of pounds of force. In a similar fashion, the thyroarytenoid muscles, with the help of forced air from the lungs, convert electrical impulses from neurons into sound energy. In the world of devices, converting electricity to sound is usually done with speakers, and converting sound to electricity is done with microphones. These instruments require diaphragms and electromagnets that are too bulky for medical ultrasound. Instead piezoelectricity is used.

Piezoelectric elements convert electrical energy into a pulse of sound wave energy and then converts the resulting sound wave echoes back into electrical energy.³ An array of piezoelectrical elements creates the sound energy needed to generate echoes, and the electrical signals elicited by their return to the probe characterizes the remarkable function.of the ultrasound transducer—the device placed on the skin to deliver and receive sound pulses. It is helpful to contrast an ultrasound transducer with the contact elements used in electrodiagnosis. EMG and nerve conduction studies (NCS) electrodes are of much simpler construction because they record and carry electrical signals without the need of converting to a different type of energy. For nerve conduction studies, there is a pair of electrodes for delivering electrical energy and a pair for recording it.⁴ These electrodes consist of exposed metal in rings, disks, or needles (Fig. 1.7). Once electrical signals are obtained, either directly in electrodiagnostic instruments, or indirectly with an ultrasound transducer, the signals are filtered and amplified by the respective instrument to generate a display for interpretation. At this point, both instruments convert electrical energy into light energy to create an image. In electrodiagnosis the display is of a single point of constant intensity that varies in vertical displacement (amplitude) over time, whereas in ultrasound the image is a 2D sea of points that vary in brightness over time. In EMG there is conversion of the display into sound energy as well. Sound is typically part of the display in ultrasound only with Doppler blood flow studies. As would be expected from their more complex design elements, ultrasound transducers are significantly more expensive and more fragile than electrodiagnostic electrodes.

Fig. 1.7 A, The devices required for performing electrodiagnostic studies: a measuring tape for measuring the distances between stimulation sites and stimulation and recording sites, tape for securing surface electrodes, a variety of surface recording electrodes (ring, bar, disk, plate), and a stimulating electrode. Two types of electromyography needle electrodes, the longer one is monopolar, the shorter is concentric; alcohol for decontaminating the skin, and electrolyte paste for reducing the impedance of the skin. B, The ultrasound transducer that is used to contact the skin of the patient, along with the gel used to enhance sound transmission through the skin. Of note, ultrasound gel works well as conducting paste for electrodiagnosis, but the reverse is not true. Not surprisingly, the set-up time for ultrasound is less than that for electrodiagnosis.

EMG textbooks often devote several pages to describing the unique features of EMG electrodes because subtle differences in structure can lead to minor, but argued over, differences in recording characteristics (see Fig. 1.7).⁴ With ultrasonography the function of the transducer is more complex than a simple recording electrode and intimately linked to the instrument, so ultrasonographers tend to fixate less on the predictable consequences of minor design variations and more on figuring out the best way to create an optimal image with the technology at hand. The following discussion of transducers serves to highlight their limitations and possibilities in evaluating patients with neuromuscular disease.

Compound Structure

An ultrasound probe actually contains an array of multiple tiny component piezoelectric transducers each providing a single line of ultrasound data (Fig. 1.8). By seamlessly stitching together each of these lines of data, the display creates a 2D image in the same way that old television screens created images from parallel lines of display.³ Ultrasound transducers come in many shapes. When using ultrasound to study the heart it helps to have a curved surface so that when the probe is placed against the ribs, it is able to scan a larger tissue area through a narrow aperture.² Internal probes for vaginal, rectal, and esophageal imaging have shapes that conform to the anatomic requirements of their use.³ At the greatest extreme, intravascular ultrasound probes can be used to examine arterial walls.⁵ All involve different arrays of transducers or a motorized transducer or set of transducers. For neuromuscular imaging, the most commonly used probe is a linear array probe because most imaging planes are compatible with this shape. A hockey stick probe, which is a linear probe with a smaller footprint and extended handle, is sometimes useful for imaging around bony surfaces, such as the medial epicondyle.

Fig. 1.8 A drawing of an ultrasound transducer that has had the impedance matching material and cover removed. Note the multiple transducer elements, each with its own electrical contact. The thickness of the ultrasound power cord (see Fig. 1.7) reflects the fact that most transducers need wiring for hundreds of elements.

Miniaturization with Piezoelectricity

The critical principle underlying ultrasound transducers is piezoelectricity, a property inherent in different types of special materials.³ This phenomenon was first described in quartz crystals. Whereas noncrystalline quartz contains a disordered arrangement of silicon dioxide molecules, quartz crystals have an orderly lattice that is determined by the electrostatic charges of the individual molecules, which incidentally have an asymmetric polar configuration. Application of a current across this lattice leads to absorption of energy, changes in electron orbits, and consequently shape changes in the underlying crystal (Fig. 1.9). This abrupt dimensional change creates a sound pulse. The application of mechanical stress to a quartz crystal, in a reverse fashion, causes an electrical current, making this material ideal for use in phonograph pickup needles or ultrasound transducers.

Fig. 1.9 This is a schematic of how piezoelectric materials respond to an electrical current. Because of the inherent asymmetry of the underlying chemical structure, when a current is applied there is a shift in molecular dipoles, causing a change in the shape of the piezoelectric element.

Currently, lead zirconate titanate,³ a piezoelectric ceramic, is used for ultrasound transducers. Although this chemical is not naturally piezoelectric, when it is heated to more than approximately 350° C in an electromagnetic field, the molecules can quickly be induced to assume a dipole-driven lattice that confers piezoelectric properties. When cooled, the application of a voltage across this material causes it to thicken or thin as the suspended molecules turn their dipoles toward or away from the electrostatic charge. This shape change, in a small wafer of material, generates the sound pulses used in modern ultrasound transducer probes. These small wafers can be arrayed in a linear fashion, each with its own separate wiring, to create the typical ultrasound transducer.

The Transducer Shell

The casing of the ultrasound probe holds a number of elements including the individual transducers, the wiring to these transducers, and several layers of material.³ Note that the cable that connects the transducer to the instrument is quite thick (see Fig. 1.4) because it contains both inputs and outputs to the hundreds of miniature transducers that make up a typical linear array transducer. Damping material is placed at the base of the transducer to help control the duration of the sound pulses emitted by the probe. Shorter sound pulses are associated with better resolution. Piano strings use dampers to enhance a staccato effect of a key strike; when these are raised by engaging the sostenuto pedal, the string vibration is prolonged, causing sounds and notes to slur together. As such, the damper material enhances the discrimination provided by pulsed sound.

Between the contact face of the probe and the underlying transducer array, matching layers are inserted.³ These help reduce the impedance mismatch of the transducer and the skin, a phenomenon further aided by the use of coupling gel. These steps enhance through-transmission of sound into the body, and without these steps more than 80% of the sound would be reflected back to the transducer at the skin surface. A more detailed discussion of acoustic impedance matching can be found elsewhere.⁶

Sound Pulse Technology

At its simplest, echolocation involves a single sound pulse, followed by an evaluation of the timing and intensity of any returning echoes.¹ Bats use their vocal cords to emit short pulses of sound to create echoes. Humans, ill-equipped for this task, can shout into a deep well and infer from the delay of the sound return something about its depth, and by the quality of the sound something about the reflecting materials in the well. If there are smooth stone sides and a water reservoir at the bottom, a sharp crisp echo will return. If instead moss covers the walls and leaves coat the bottom, a muffled softer sound will be heard.

Ultrasound relies also on single sound pulses to create an echo trace.³ The sound is produced by the transducer element, the sound penetrates soft tissues, and as the sound passes through the body, tissue interfaces at different levels create multiple echoes of different intensities. At each interface, some sound is reflected and some is transmitted. The echoes return at different times to the transducer element, which converts them to a trace signal. The latencies of the echoes provide information about spatial relations (depth of the reflective surface), and the intensity of the echoes provides information about the types of reflecting surfaces.

Humans primarily use sight to map space, and this is why ultrasound data are displayed as an image. Most vision is based on a source that bathes a scene with a constant flow of light energy, such as the sun or a light bulb. This contrasts with ultrasound, which maps space with pulsed sound; its equivalent would be a fast-pulsing strobe light. An echo is a reflection of an entire sound pulse, so the shorter the pulse, the less likely an echo will return during the pulse itself, and the less likely that early and late components of sequential pulses will interfere with each other.³ Typically a B-mode ultrasound pulse is 2 to 3 cycles of sound energy in duration, and at 10 MHz, this takes about 0.2 to 0.3 μs. Of interest, bat calls are of much longer duration: 0.2 to 100 μs, but their calls cover much greater distances, perhaps 50 meters or more. Appropriately, call duration shortens, improving localization, as a bat approaches a target of culinary interest, such as a moth.

Ultrasound transducers emit many pulses of sound per second, which constitutes the pulse repetition frequency.³ It can be thought of as similar to the frame rate on a video camera, and it is of importance in determining the temporal resolution in an image. Duty factor is a term that calculates the percentage of time the transducer is actually emitting pulses, which is substantially shorter than the time it spends receiving echoes. The total energy emitted by a transducer correlates directly with the duty factor and determines certain physical effects such as tissue heating.

Tissue responds to insonated sound pulses by generating echoes. It is the job of the instrument to re-create an image from this information. How the instrument does this requires an understanding of how pulsed sound behaves in human tissue.

Sound: Fundamental Principles and Its Behavior in Tissue

Speed of Sound

Sound is a traveling pressure wave. Unlike an ocean wave, in which the direction of the amplitude (vertical) is out of plane with the direction of travel (horizontal), sound waves consist of alternating compressions and rarefactions of molecules in the medium through which it travels (Fig. 1.10). Such changes can be measured by pressure, density, or particle vibration, all of which can be considered acoustic variables.³ As with any wave, sound has intrinsic detectable wavelengths, periods, and frequencies, all of which are interrelated. The speed of sound propagation, however, is not a property of the type of sound energy, but rather of the medium of transmission. Sound propagation speed is related to stiffness of a tissue, or more precisely, hardness. As such, it travels slowly in easily compressed gases, more rapidly in fluids and body tissues, and fastest in solids. An intuitive sense of sound propagation is readily derived from experience with a Slinky, a toy originally constructed of a wound coil of flexible metal (Fig. 1.11). Mechanical energy creates standing waves of alternating rarefactions and compressions of successive spirals of the coil; these waves travel faster if the coil is tightly wound with harder metal. Tightly wound Slinky toys travel down steps faster than less tightly wound and looser versions (e.g., plastic).

Fig. 1.10 This is a schematic comparing a traveling wave (the superimposed sine wave) with the alternating compression and rarefaction of molecules that are seen with the compression wave of sound energy.

Fig. 1.11 This is another way to compare a transverse and compression wave; the bottom figure resembles that of a Slinky toy.

The variation of sound propagation in different materials is of significance both generally and clinically. An approaching train can be detected sooner by putting one’s ear to the rail than by listening through air for the sound it makes. On average, sound travels about four times faster through metal than through human soft tissue, which is minimally faster than through sea water, and in turn about four times faster than in air. The hardest substance known to man, diamond, also conducts sound the fastest at 12,000 m/s (Table 1.1).

Table 1.1 Speed of Sounds in Different Substances

Material	Velocity (m/s)
Air	331
Fat	1450
Water (50° C)	1540
Human soft tissue	1540
Brain	1541
Liver	1549
Kidney	1561
Blood	1570
Muscle	1585
Lens of eye	1620
Tendon	1650
Ice	3152
Skull bone	4080
Brass	4490
Aluminum	6400
Diamond	12,000

Ultrasound Range Finding

A key goal of ultrasound is an accurate representation of spatial characteristics of human tissues. To measure depth (range), ultrasound relies on the timing of the return echoes.³ As mentioned above, ultrasound makes a number of assumptions to calculate depth. It assumes that sound travels in a straight line from the transducer to the target and back, that enough echoes are reflected back to the transducer from relevant tissue structures to be captured and measured, and that the speed of sound in soft tissue is sufficiently consistent across tissues to generate an accurate measurement. In general these observations hold true, and particularly for the more superficial structures studied in neuromuscular ultrasound. In a fashion similar to an F-wave when performing electrodiagnostic studies,⁴ an echo reflects a round trip of sound from the transducer to the structure of interest, and back again to the transducer. All ultrasound instruments assume that sound travels through tissues at a constant rate of 1540 m/s, and images from echoes are constructed based on this calculation. It may be easier to understand this number expressed differently: for each centimeter of depth, the average round trip travel time for an echo in human soft tissue is 13 μs.³

The width on an ultrasound image represents the summed thickness of the individual transducer elements within a linear ultrasound probe (see Figs. 1.4 and 1.8).³ In some ways width measures in tissue may be slightly more accurate than depth measurements in that they do not assume constant sound speed in tissue. However, because width measures do not take refraction of sound into account, there can be distortion of width as sound passes through tissue layers of different sound transmission speeds, if these are oblique to the angle of insonation (see oblique echoes later).

For practical purposes, the variation of sound speed is rarely important. However, if someone is measuring the depth to the sacrum in two patients, one with minimal body fat and one with 7.5 cm of fat over the gluteus maximus (not unusual in obese patients as measured by EMG needle length), there would be a subtle effect. Ultrasound travels slower through fat than through the “average” human tissue, and as such, ultrasound tends to overestimate the thickness of a fat layer by about 6% (1541/1450), and underestimate the thickness of muscle by about 3% (1541/1585) (see Table 1.1).

It would be tempting to speculate that the speed of sound in nerves would be somewhat less than that of average human tissue because of its high fat content, but at this time this value is unknown. However, the speed of ultrasound in brain is 1541 m/s, and most nerves are sufficiently thin as to make any special adjustments for size moot. Special techniques, using ultrasound radiofrequency analysis for measuring body tissue and fat thickness may be of value for addressing questions of this type.⁷

Discrepancies in sound transmission speed affect area and volume measurements more than linear measurements because area is related to the square, and volume the cube, of the linear measurement. This may account for some of the slight variation in estimated thickness of muscle or fat by ultrasound and by other imaging modalities such as MRI or CT. At this time, standard volume measurements are not routinely calculated for muscle or nerve, but they are for internal organs based on a set of standard linear measures from 2D ultrasound. The kidneys,⁸ for example, use the product of 0.5{A-P diameter} × {transverse diameter} × {height} to produce a fairly reliable estimate of kidney volume. Adjustments for variations in speed of sound in neuromuscular tissue could be incorporated into mathematical models of this type when it becomes routine to calculate volumetric measurements.

Another factor that influences sound speed in tissue is temperature. This is most striking for fat—the warmer the temperature, the slower the sound speed—but the reverse holds true in other tissues, although to a lesser degree.⁹ In general, temperature effects are of limited significance during routine imaging. Of interest, there is some evidence that fat in humans simultaneously exists in solid, liquid, and melting phases, perhaps with a key transitional phase occurring at 35° C.¹⁰

The Relationship Between Propagation and Harmonics

In practice, the propagation speed of a wave is not even. As the wave travels through any medium, its higher pressure component travels faster than its lower pressure component, so its progress through space is somewhat nonlinear.^3,⁹ As a result, the wave acquires superimposed higher frequency harmonic components that increase over the distance covered (Fig. 1.12). Harmonic frequencies consist of multiples of the original or fundamental frequency of the sound. The concept is of clinical relevance in situations in which the primary frequency creates problematic artifacts. For example, sometimes when imaging deeper structures through ribs, there is less artifact with harmonic frequencies, and analyzing return echoes only at these higher frequencies may enhance the imaging of deep structures.

Fig. 1.12 In nonlinear propagation, propagation speed depends on pressure. A, Higher-pressure portions of the wave travel faster than the lower-pressure portions. B and C, The wave changes shape as it travels. This change from the initial sinusoidal shape introduces harmonics that are even and odd multiples of the fundamental frequency.

Attenuation

Attenuation refers to the reduction of sound intensity as it passes through tissue. Some of this energy is lost through reflection or backscatter, but the majority is lost through absorption of the sound energy.^3,⁹ The degree of attenuation can be measured in several ways and is expressed as decibels (dB) of sound energy attenuated over distance (dB/cm). A number of factors influence attenuation and backscatter. For example, attenuation is frequency dependent, such that higher sound frequencies are dissipated through distance much quicker than lower frequencies (Table 1.2).³ Sounds in the audible range show similar properties, which is why blaring music in a passing car is heard primarily in lower-pitched base tones, and why thunder from distant lightning is a low-pitched rumble, whereas a nearby lightning strike has a high-pitched and ominous crack. The most critical principle explained by this phenomenon is that higher-frequency ultrasound has a lower effective depth of penetration for imaging than lower frequency.^2,^3,⁹

Table 1.2 Average Attenuation Coefficients in Tissue

Each body tissue attenuates sound to a different degree (Table 1.3). The relationship between frequency and absorption in a given tissue is typically linear over standard ultrasound frequencies, with an increase in attenuation with higher frequencies. The slope of this effect, however, varies between different tissues ^10,¹¹ and across species. In fact, the attenuation coefficient itself may not be entirely uniform for the same tissue in a different location in the same individual. For some tissues, such as tendon and skeletal and cardiac muscle, the attenuation coefficient nearly doubles if the tissue is imaged in a plane parallel to the direction of muscle (tendon) fibers as compared with a cross-sectional image.¹² One might imagine how sound energy might be redistributed more easily among cellular molecules from insonating a small cross section of hundreds of muscle fibers, than insonating a few muscle fibers along their entire length.

Table 1.3 Attenuation Coefficients of Biologic Substances (at 1 MHz)

Material	α(dB / cm)
Water	0.0022
Amniotic fluid	0.0053
4.5% albumin	0.019
Blood	0.18
Fat	0.63
Brain	0.85
Liver	0.94
Kidney	1.0
Bone	20
Lung	41

The phenomenon of absorption is critical for therapeutic applications of ultrasound. As sound energy is absorbed by tissue, it is converted to heat. Ultrasound therapy was introduced to the specialty of physical medicine and rehabilitation many years ago, specifically for this purpose. Because sound energy is more readily absorbed by tendons, ligaments, and bones, it is a means of providing direct heating to deep structures that are vulnerable to common musculoskeletal injuries ^3,¹³ and often the last to be warmed by surface heating. (Refer to other sources for further information about this type of therapeutic ultrasound.¹⁴^–¹⁹) However, it should be noted that with increasing amounts of insonated energy, additional safety precautions are needed.²⁰

One of the most interesting theoretic applications of ultrasound energy is the ability to activate microencapsulated drug release with an ultrasound beam.²¹^–²⁵ Microencapsulated drugs injected intravenously circulate throughout the entire body, but the capsules release the drug only in tissues being actively insonated, thereby significantly enhancing concentration at the target site. This approach can be used to quite selectively deliver drug to a tumor, for example, while minimizing its distribution in other healthier tissues. Other interventional applications of insonation and local microcavitation, such as enhanced healing of bone injury²⁶^–²⁹ or as a tool to possibly facilitate muscle regeneration are under active study.²⁹

Echo Behavior

Perpendicular Echoes and Acoustic Impedance

As sound energy courses through tissue it creates echoes. The simplest scenario to consider is one that involves a perpendicular incidence of the direction of the ultrasound sound wave to a smooth interface of two layers of tissues (Fig. 1.13). A significant portion of sound energy is transmitted and continues in the same direction. However, the reflected energy at this interface reflects back in the opposite direction directly toward the transducer.³ The amount of reflected energy is proportional to the difference of acoustic impedance of the two layers; the greater the difference, the greater the reflected energy. When looking through water at a jellyfish, the change in acoustic impedance between air and the surface of the water is much greater than that between water and the surface of the jellyfish. As such, the air-water boundary is much easier to discern than the water-jellyfish boundary, because more light is reflected at the former. The ratio of reflected to incident energy is the intensity reflection coefficient. The ratio of the transmitted energy to the incident is the intensity transmission coefficient. The sum of these two ratios is one (an assumption that assumes that there is very little if any absorption of sound energy at that particular barrier). Acoustic impedance, measured in rayls, relates to the speed of transmission of ultrasound in a particular tissue or layer. More thorough discussions of the complex topic of acoustic impedance, both characteristic and specific, can be found elsewhere.^3,³⁰^–³¹

Fig. 1.13 Demonstration of sound passing through a door. Note that the sum of the transmitted and reflected sound equals the intensity of the incident sound.

Oblique Echoes

If the angle of incident sound is oblique, and the boundary is smooth, the angle of incidence and angle of reflection (compared to perpendicular) are equal—a phenomenon analogous to a pool ball ricochet.^3,³⁰^–³¹ The angle of transmission is also related to acoustic impedance. If the deeper layer conducts sound faster than the superficial layer, the sound bends toward the boundary, and if it is slower, sound bends toward the perpendicular. These are examples of refraction (Fig. 1.14).

Fig. 1.14 This figure represents the transmission and reflection of sound energy from air to glass, when the incident sound is oblique to the reflecting material. Note that θi equals θr, and that θt is determined by the relative ratio of the acoustic impedance of air to glass.

Scattering

Of course, smooth boundaries are not the rule in biologic tissues. Rough boundaries and heterogeneous materials tend to cause multidirectional transmission of incident sound (scattering) and reflection (backscattering).^3,³⁰^–³¹ Most light reflectors are not smooth; a flashlight aimed at a wall can provide a diffuse dim light for a room because of its backscatter (Fig. 1.15). Fog, an example of heterogeneous materials in air, causes both scattering of transmitted light and backscattering of reflected light, which is why headlights set on high beam in fog are not the advantage they are in clear weather. Speckle is a consequence of the random backscattering of sound pulses in an imaged plane of tissue. These waves can interfere positively or negatively with each other or with waveforms that have been reflected perpendicularly back to the transducer, causing random decrements and enhancements of brightness, or speckle.

Fig. 1.15 This figure shows how structures with no (A), mild (B, as in nerve), marked (C, as in tendon), and complete (D) anisotropy reflect incident energy, showing an inverse relationship of backscattering to inherent anisotropy (structures with high anisotropy have little backscatter; see Fig. 1.23 for comparison).

Contrast Agents

Ultrasound contrast agents are examples of small heterogeneous materials that can be injected into tissues and cause a local increase in the echogenicity of tissue vasculature with multiple backscattered echoes. Unlike radiographic contrast agents (intravenous pyelography dye) or MRI contrast agents (gadolinium) that work in solution and can demonstrate leakage of contrast into perfused tissues, the echogenicity of ultrasound contrast agents is based on their reflective properties, which means that they must be particulate, too large to cross capillary gaps, and too small and too pliable to obstruct blood flow. Typically these consist of microbubbles of gas encapsulated within a lipid or protein.^22,^32,³³ As such, they can demonstrate areas of increased blood flow within tissue, but not areas of capillary leakage. Of interest, unlike other contrast agents, intravascular adhesion and retention of microbubbles may occur in areas of vascular wall inflammation.^32,³³ Thus vascular wall changes that lead to fibrinogen or platelet adhesion and clot formation may be more sensitively detected by certain types of ultrasound contrast agents than by soluble contrast materials used in MRI and CT. It is possible that ultrasound contrast may find new applications in neuromuscular disorders such as inflammatory muscle or nerve disease.

Focusing

Ultrasound beams can be focused in one dimension to minimize beam width (the width of the sound out of plane with the image created).^3,^30,³¹ This type of focus is different than optical focusing, which occurs in three dimensions, a property readily demonstrated with paper, a magnifying glass, and bright sunlight. The end result of ultrasound focusing, like optical focusing, is enhanced resolution. By narrowing the width of the sound, there is a reduction in the out-of-plane averaging that takes place within an image, and therefore resolution is improved. For imaging focal swelling in a nerve in a cross-sectional image, this property is of value because it allows for more precise identification of the point of maximal enlargement of the cross-sectional area of the nerve. Focusing can be done by shaping the array in a convex fashion, phasing the sound wave pulses of the individual elements in the transducer and by careful study of the natural focusing that follows any sound beam. Multiple focal zones can be obtained in some images, but this requires that different pulses are used for different focal distances, which therefore reduces the frame rate (see temporal resolution).

For the ultrasonographer, the optimal focus of the ultrasound beam is often illustrated on screen by a small arrow to the right of the image (see Fig. 1.6). It points to the depth where the image quality is typically optimal, and can be adjusted in terms of depth and in terms of how thick the band of focus is. When measuring small structures, such as nerves, it is important to optimize the focus to the appropriate depth. For example, when following the median nerve from the wrist through the forearm, several changes in focal depth are necessary to optimize the image of the nerve as it passes under the flexor digitorum superficialis and then the pronator teres.³⁴ Novice clinicians frequently forget to adjust the focal zone, an error that makes following a nerve as it approximates the surface or dives beneath muscles more challenging. More detailed discussions of probe design and focusing are available elsewhere.^3,^30,³¹

Resolution

The goal of ultrasound is optimizing resolution of structures of interest. An objective test of this ability can be provided by a phantom, which is conceptually similar to an eye chart. A phantom is any type of tissue model that can be scanned.³ Such items often are agar filled, with suspended smaller objects that simulate tissue structures that can be readily imaged with ultrasound, and different designs are available for purchase. The ability to resolve such smaller items (either objects in the phantom or letters on an eye chart) is a measure of resolution. A phantom can be used to compare resolution of different transducers and different instruments.

In ultrasound, axial resolution refers to the ability to distinguish nearby objects in the direction of sound travel; lateral resolution is in the direction perpendicular to this (which would be across scan lines). Axial resolution is related primarily to the pulse length of the emitted signal, which is frequency dependent, and the lateral resolution relates to beam width, which relates more to focusing. Therefore, when evaluating different instruments, it is critical to look beyond a single technical reference (such as transducer frequency), and to compare image quality in tissues of interest.

Temporal resolution in ultrasound refers to the ability of ultrasound to discriminate quick motions in tissue.³ Higher frame rates allow for better imaging of motion and smoother definition of it. Higher temporal resolution involves narrower focuses and reduction of persistence and averaging of images. The latter provide a means of controlling random artifacts such as speckle, but tend to reduce temporal resolution. Because the quality of a static image is easily compared with MRI or CT, there is a bias to design ultrasound instruments to maximize static image quality. Interventional/vascular radiologists and clinical subspecialists have provided much of the impetus for real-time motion and blood flow detection. Given the complexity and flexibility of ultrasound instrumentation, enhancements in temporal resolution are not only possible but may be critical for studying subtle quick movements, such as fibrillation potentials in muscle,³⁵ and adaptive instrument design may be of future value.

Ultrasound Instrument Function: Driving the Transducer and Image Display

The ultrasound machine carries out a number of important tasks, some of which have been discussed already, such as sound pulse frequency and phasing and image display. The complexity or accuracy of these functions is not dependent on the size of the instrument nor its display screen, but rather the sophistication of its underlying technology. Some instruments are now run by laptop computers and provide competitive image resolution. Ultrasound instrumentation can be broken down into efferent and afferent functions, which will be described in sequence.

Efferent Ultrasound Functions: Beam Formers, Pulsers, Coded Excitation, Amplifiers, and Transmit/Receive Functions

The efferent functions of ultrasound instruments are simpler than its afferent functions and begin with the beam former.^3,^33,³⁴ The generation of a sound pulse requires several devices. The first is a pulser that takes power line electricity and translates it into brief pulses of voltage. The device is not dissimilar to the stimulator on an EMG machine, which delivers single or multiple timed pulses of voltage or amperage of a set duration. In the ultrasound instrument, however, this is somewhat more complex. Ultrasound probes consist of arrays of piezoelectric transducer elements, and each element receives a separate input from the pulser (one per scan line). As such, an ultrasound instrument’s pulser operates hundreds of channels, rather than the single or dual stimulator systems found in most electrodiagnostic instruments (see Fig. 1.8). To further complicate the process, for the purposes of multiple focusing and image enhancement there are often coded timing variations (coded excitation) in the order in which the transducers are activated. Thus a series of pulse delay switches are automatically built into the instrument.

In tandem with the pulser is the amplifier, which determines the intensity of the insonated sound. Once the pulse sequences are ready for relaying to the transducer, there is a final gate. This is a transmit/receive switch in the instrument that following each pulse sequence automatically switches off the emit (efferent) function of the transducer and turns on its receive (afferent) function, so it can analyze returning echoes. The transducer is in the receive mode the vast majority of the time—typically 99.0% to 99.9% of the time—although more emit time is required for Doppler imaging. Pulse duration, sequencing, and on/off time cycles are preset based on the type of imaging (Doppler versus B-mode), transducer type, focusing, and imaging depth used. Direct operator control is usually restricted, however, to only modulating the amplitude of the emitted signal.

Afferent Functions of the Instrument

Once the tissue has been insonated with a single pulse, the instrument switches to receive and responds to the returning echoes. The instrument then switches back to delivering sound pulses. This cycle is repeated continuously during imaging.

Amplifiers and Time Gain Compensation

After the piezoelectric elements convert returning echoes into electrical impulses, they need to be amplified to create a display. Most instruments also contain a time gain compensation panel to modify the process when needed by the ultrasonographer. The echogenicity of unmodified distant echoes is much less than that of near echoes. This is because of the attenuation of ultrasound energy as it passes through tissue, both in transmission to the target and in reflection back to the transducer. As such, instruments are all designed to amplify later echoes based on a standard estimate of the degree of soft tissue attenuation (see Tables 1.1 and 1.2). However, some tissues cause more or less absorption of sound than the predetermined average, and in these cases, prudent adjustment of near and far gain on the time gain compensation panel can result in a more representative image of the structures of interest.³ Alternatively, adjustments are sometimes helpful for reducing noise from certain irrelevant layers in an ultrasound image. The time gain compensation panel (see Fig. 1.2), has a characteristic display control, and it controls the degree of amplification at different depth layers. Adjustments to this panel are most helpful when imaging through tissues that excessively attenuate sound, such as fibrous tissue, or fail to attenuate it, like cystic or vascular structures (Figs. 1.16 and 1.17).

Fig. 1.16 These two figures provide examples of shadowing. A, B-mode cross section of the anterior tibialis about halfway between the knee and ankle. The two images show the same data, on the left displayed in color (white being brightest, then green, then red, then blue being darkest). On the right is a standard gray-scale image. The prominent, echogenic, and collagen-dense central aponeurosis of this muscle reflects a greater proportion of incident sound than expected and as such, the muscle deep to this layer, particularly in the lower right portion of the image where the membrane is thicker is less echogenic (or more blue). On the upper left, where there is no aponeurosis, the muscle is somewhat hyperechoic (green). These changes are subtle, but are of some importance if analyzing tissue echogenicity. Color display is not routinely used, but in this case highlights the subtle shadowing associated with an aponeurosis. The bone edge of the tibia completely reflects sound, so the bone substance appears as black in both images.

B, A cross section through the deltoid in a patient with post-traumatic myositis ossificans. The calcification in muscle (arrow) causes significant shadowing (S) in some areas shadowing out the otherwise prominent presence of the bone edge of the humerus (H).

Fig. 1.17 This figure provides an example of enhancement. A cross-sectional image of the medial gastrocnemius. On the right, the image is taken with only minimal pressure applied to the transducer, and a large vein is present. Deep to this is an area that is relatively hyperechoic (arrows). On the left, the image is taken with increased pressure, the vein is compressed and the hyperechoic area disappears, indicating that this is an enhancement artifact secondary to the presence of relatively hypoechoic blood in the overlying venous structure.

Filters

Ultrasound instruments also use filters to enhance images. As mentioned, attenuation is frequency dependent, with higher frequency signals being disproportionately attenuated as sound traverses a distance. Also, sounds emitted by a transducer have a modest range or bandwidth of frequencies. As echoes return from deeper structures in tissue, the highest-frequency sounds are attenuated more than the lowest-frequency sounds in this bandwidth. As such, the ratio of high frequency noise to high-frequency signal is increased in deeper tissues. Many instruments selectively filter out more high-frequency sounds from the bandwidth for returning echoes of increasing latency, such as the ones from deeper layers.³ This graded use of the filter with depth improves the signal-to-noise ratio. Harmonic imaging can also be useful. As indicated, sometimes harmonics of the insonated frequency are less subject to distortion than the fundamental frequency, and this can be used to advantage in certain types of ultrasound imaging.

Persistence

Persistence is a form of image averaging over time in ultrasound that improves spatial resolution. This removes random speckle (the unpredictable echo changes that result from the positive or negative interference of backscattered sound), and other artifacts. However, as persistence increases, temporal resolution decreases.³ Instruments that appear to be ideal because of simplified panel configurations have more preset and unalterable imaging features. Such preset configurations may emphasize persistence and stability of an image, and therefore temporal resolution may be sacrificed. If the need of an instrument is to detect rapid tissue or needle movement, more setting flexibility may be helpful.

Amplitude (A)-Mode Imaging

Both EMG and ultrasound began with the use of oscilloscopes ^3,⁴ (Fig. 1.18) using a standard A-mode display. In A-mode a single line of ultrasound data only is displayed, with the X axis representing echo latency (which corresponds to distance or depth) and the Y axis representing echogenicity (e.g., echointensity amplitude). This should be familiar because it is the typical display for EMG or NCS, where the X axis is latency and the Y axis is amplitude of the electrical signal (voltage). For an electromyographer it can be puzzling to understand how ultrasound evolved into a real-time 2D image, while EMG has remained as a single real-time linear display. With ultrasound, a change in amplitude comes from an echo, which is generated by a fixed anatomic structure in a distinct spatial and temporal relation to the transducer’s incident sound pulse. With electrodiagnosis, a change in amplitude comes from measurable progressive ion movements along the length of excitable tissues within any radius of the recording electrodes. Although it is possible to triangulate bioelectrical signals and to localize them accordingly, this is complex and not routinely performed with electrodiagnostic instruments. The key difference is that in ultrasound, amplitude variations (echoes) occur in a temporal pattern determined by the location of the reflecting anatomic structure, which immediately lends itself to spatial display. In electrodiagnosis, spatial relationships can only be inferred.

Fig. 1.18 These are photographs of oscilloscopes used in original ultrasound research at Wake Forest University in the late 1960s. A, This image shows how images were captured using a Polaroid camera. B, This image shows the traditional A-mode display familiar to most electromyographers.

Brightness (B)-Mode Imaging

One alternative to A-mode imaging is B-mode, or brightness mode imaging. In this case, instead of displaying amplitude by variations in signal amplitude on the Y axis, amplitude is represented by varying the brightness of the trace. Higher-amplitude signals are brighter than lower-amplitude signals. The audio recording of an EMG signal parallels B-mode imaging in that the louder the sound of the sound of the EMG signal, the greater its amplitude. Because time, in ultrasound, is equivalent to distance, if one uses brightness to represent amplitude it opens up a display screen for multiple parallel A-mode recordings. As a composite, this allows varying degrees of brightness to represent anatomic structures within a 2D map of width and depth. However, this type of display does have some minor limitations.

Compression

Displaying amplitude as brightness instead of vertical displacement is somewhat more problematic than one might expect. The human eye can readily distinguish subtle variations in height but distinguishes variations in light intensity poorly by comparison. Further, although there is a direct linear relationship between measured and perceived height, the relationship between measured and perceived brightness is exponential ³⁶ and intuitively much less clear. A simple experiment helps clarify this. In the dark, a single flashlight has a substantial effect on the ability to see, an effect that is far from doubled by the addition of a second flashlight. In fact, it is not clear exactly how many flashlights would be needed to double the perceived intensity of ambient light. As a result, to represent brightness on a display screen in a proportionate way, it cannot simply be linear; it must be an exponential increase. Furthermore, the increase needs to be adjusted to compensate for any pupillary constriction that accompanied the increased brightness. Not surprisingly, there are no simple formulas for doing this in such a way as to provide the same intuitive sense of proportion of height and amplitude as in A-mode. Of interest, the formulas used for determining how amplitude is translated into brightness vary from manufacturer to manufacturer and instrument to instrument, and are often empirically determined by image aesthetics of select tissues of the manufacturer’s choice.

A-mode is commonly used in ultrasound of the eye and orbit.³⁷ The technique provides a more quantitative approach to assessing echogenicity. It not only provides an accurate assessment of reflective strength from structures within the eye but also a means of comparing reflective strength from a known ocular structure (such as the posterior scleral wall) and that of an unknown mass within the eye. Unlike B-mode gray scale parameters, A-mode amplitude to decibel ratios (compression metrics) are standardized across different instruments and manufacturers in ophthalmology ultrasound devices. There remain problems, however, with ceiling effects of the display as apparent in Figure 1.19. At this time, A-mode is not routinely available on most commonly used nonophthalmology instruments.

Fig. 1.19 B-mode (A) and A-mode (B) images of the orbit. The A-mode image is from a horizontal beam in the center of the image. Note that A-mode records amplitude variations in only one linear direction and that the amplitudes correspond to variations in brightness. It is difficult, however, to interpret the echogenicities displayed because the peak amplitude of the signals cannot be determined as a result of a limited range of display mode (ceiling effect). Also note the substantially greater information conveyed by B-mode.

(From Waldron RG, Aaberg TM: B-scan ocular ultrasound, eMedicine from webMD. Updated January 2009. Available at www.emedicine.medscape.com/article/1228865-overview.)

M-Mode Imaging

M-mode is useful for measuring latency changes in moving tissue. This technique depends on selecting a single channel of information from B-mode ultrasound and displaying how it changes over time (Fig. 1.20).³ This display is more like an electrocardiogram (ECG) tracing, with the time base measured in tenths of seconds. This technique can be used to measure muscle contraction or fasciculation contraction time.³⁵ It should be noted that electrodiagnostic techniques measure the duration of the muscle action potential, which is a surface membrane–only event lasting on the order of 10 ms, whereas ultrasound measures actual mechanical contraction and relaxation time, on the order of 200 to 300 ms.

Fig. 1.20 A, A cross-sectional B-mode image of the tibialis anterior midway between the knee and ankle. The thin green vertical line in this image is used to create the M-mode display below which shows changes in this single line as the muscle changes over time. The plus signs mark the onset of a compound muscle action potential created by supramaximal stimulation of the peroneal nerve at the fibular head. The duration of the actual mechanical contraction of the muscle is 241 ms, which is much longer than the compound muscle action potential duration, and only reflects the timing of electrical changes across the membrane. Also note that the peak displacement of the muscle (peak contraction time) occurs about 71ms after onset.

B, The compound muscle action potential of the same muscle recorded with stimulation of the peroneal nerve at the fibular head and surface electrodes straddling the plane of this image. Note the striking difference in duration; by ultrasound the duration is 241 ms; by nerve conduction studies it is 10 ms, demonstrating the difference between the mechanical and electrical behavior of muscle.

Digital Conversion

All modern ultrasound instruments use analog digital conversion to display signals. This permits averaging for persistence as well as for interpolation or smoothing.³ Such adjustments can affect perceived echogenicity. Some instruments provide user control of these functions, and adjusting these controls may help enhance imaging of different types of structures.

3D and 4D Imaging

With computer-based technology, 2D images can be reconstructed into 3D and 4D images. The technique is at its most dramatic when creating realistic videos of fetal facial expressions and movement. The presence of a face surrounded by amniotic fluid is an ideal environment for the technology (Fig. 1.21).³ In clinical practice, however, most ultrasound imaging is still done in 2D planes (although technically this is really 3D imaging—depth, width, and time). Video is not routinely included in most published ultrasound papers and 3D images are relatively uncommon. As such, readers of the literature do not routinely find such images other than as online supplements.³⁸ At this time little has been published on the use of 3D imaging in nerve or muscle.

Fig. 1.21 This is an example of a 3-dimensional (3D) image now routinely available for depicting the detailed surface anatomy of a fetus. In practice, 3D imaging is most informative for structures surrounded by fluid or involving fluid filled structures. The technology is still evolving.

(From Wladimiroff JW, Eik-Nes SH, editors: Ultrasound in Obstetrics and Gynaecology, Edinburgh, UK, Elsevier.)

Color Doppler

Many ultrasound instruments have the capacity to demonstrate tissue movement and blood flow using the Doppler effect.³ The Doppler effect is a change of the frequency of sound energy by a moving object, familiar as the increasing pitch of a train whistle as a train approaches (and its decreasing frequency as it recedes). Returning echoes demonstrate a similar effect, and Doppler principles are used to generate the timing mechanisms of automatic sliding glass door openers.

The physics of ultrasound Doppler are complex and its details are beyond the scope of this text because at this time few papers have been published on blood flow or tissue motion in nerve or muscle. Nonetheless, this is an area of likely growth in the field of neuromuscular ultrasound. Initially much of the focus of color Doppler was on the detection of abnormal vascular blood flow, for the purpose of identifying critical occlusion or pending occlusion of major vessels. For the purpose of neuromuscular ultrasound, it turns out that most of the information is coded by 2D dimensional color-flow Doppler or Doppler power displays. When examining tissue, such as nerve, for increased blood flow, it is often helpful to use power Doppler imaging. This type of imaging is less subject to noise contamination, and is better for detecting slow flow and looking at smaller or deeper vessels than spectral Doppler imaging. The results of Doppler power displays are mainly qualitative and may vary somewhat across instruments. Nonetheless, these displays provide a simple and reliable means for assessing for the presence or absence of changes in tissue vascularity. Of interest, ultrasound color allows for information coding in terms of hue and saturation in addition to brightness (luminance), and color displays therefore allow for coding of useful information not available from simple gray scale imaging. Future Doppler technology may assist in the assessment of muscle contraction and nerve movement. Readers wishing to explore color Doppler imaging in greater detail may find Frederick Kremkau’s textbook a useful starting point.³

Image Storage and Manipulation

Many modern instruments can store a final image and allow postprocessing manipulations.³ At their simplest these include annotation (arrows and text) and distance measurements, but many instruments can further alter the display by changing other features such as gray scale and compression. Such properties are useful for record keeping and research.

Quantifying Ultrasound Findings

Spatial Dimensions of Tissue

Ultrasound instruments are ideal for measuring spatial variables such as thickness, width, depth, area, and volume. Edge-to-edge resolution with ultrasound is more precise than that obtained with CT or MRI, and all instruments are designed to make measuring spatial parameters simple and quick. In practical use a few simple linear measures of structures such as the kidneys can be used to estimate kidney volume based on formulas established in cadaveric studies,⁸ but similar formulas have yet to be developed for muscle or nerve. A few caveats are warranted with linear measures: The first is that depth measurements are based on the speed of sound in tissue, and this varies slightly among different types of tissue. Width measurements depend on the assumption that ultrasound waves are transmitted through tissue in a straight line, yet refraction does occur when sound crosses planes of tissue with different acoustic impedance (such as muscle and fat). In calculating volumes, any linear errors of measurement are magnified by multiplying width, length, and depth, which should be corrected for if comparing ultrasound volume measurements with CT or MRI. Measurement error in ultrasound therefore is minor and, to some extent, predictable. Understanding ultrasound physics can be useful in addressing such minor errors if clinically indicated.

Quantifying Tissue Movement

Although ultrasound provides accurate estimates of the movement of blood in the cardiovascular system readily and quickly,³ solid tissue movement is not easily calculated³⁹ This does not necessarily imply that tissue movement is more difficult to measure than blood movement, but rather reflects the years of engineering invested in ultrasound technology for studying hemodynamics. Although ultrasound-based calculations of blood flow are not exact measurements they are reliable enough surrogates to be used routinely in clinical decision making. Nerves and muscles move, including fasciculations, fibrillations, nerve subluxation, and nerve sliding, and standard ways to measure these movements have yet to be developed. Descriptive rating scales are occasionally used and can be reliable even if they are not fully validated. Research in this area is progressing, however, and in the future more quantitative measures of muscle and nerve movement may become standardized using technology such as color tissue Doppler imaging.

Quantifying Echogenicity

Since Heckmatt and Dubowitz ⁴⁰ described the increased echogenicity of affected muscles in Duchennes muscular dystrophy 30 years ago, investigators have been intent on finding ways to measure echogenicity. Despite the substantial improvements in instrumentation in these three decades, quantifying echogenicity remains problematic. The reasons for this relate to the tension between the use of ultrasound to optimize image aesthetics and the use of ultrasound to measure physical parameters of insonated tissue. Antiglare sunglasses are a useful analogy. These reduce apperception of total light intensity (glare) in one’s environment for the benefit of enhancing the ability to see objects within it; a trade-off most users feel is worthwhile. A similar trade-off has been made with current ultrasound instrumentation, but such an approach has drawbacks.

Ultrasound instruments are designed to maximize tissue recognition and diagnosis. Recognition is defined as an awareness that something perceived has been perceived before (e.g., recognition). This definition is similar to what we typically mean by diagnosis, in which it is implied that we have seen a similar disease pattern before. Although it would seem that quantifying the physical aspects of one’s environment would go hand in hand with recognizing it or perceiving it, in fact, they are by no means equivalent. This distinction is a focus of an entire field, Gestalt psychology, which explores the nature of recognition. In the world of human experience, quantifiable aspects of sensation are separate from the more subjective aspects of perception:

The variables of sensory discrimination are radically different from the variables of perceptual discrimination. The former are said to be dimensions like quality, intensity, extensity and duration; dimensions of hue, brightness and saturation, of pitch, loudness, and timbre, of pressure, warm, cold and pain. The latter are dimensions of the environment, the variables of events and those of surfaces, places, objects, of other animals and even of symbols. Perception involves meaning: sensation does not. To see a patch of color is not to see an object, nor is seeing the form of a color the same as seeing the shape of an object. To see a darker patch is not to see a shadow on a surface…to have a salty taste is not to taste salt…to feel a local pain is not to feel the pricking of a needle, to feel warmth on one’s skin is not to feel the sun on one’s skin.

—J. J. Gibson ⁴¹

Ultrasound instruments have been designed to maximize perceptual discrimination, not pure sensory discrimination. This process, by design, sacrifices accurate representation of tissue echogenicity if it detracts from accurate perception. In like manner, human eyes sacrifice brightness discrimination by the design of the pupillary reflex, which constricts or dilates the pupil to minimize the distracting effects of changes in ambient illumination. Further, some quantifiable visual variables are of greater importance than others in daily life. For example, size discrimination is, in general, of greater importance than brightness discrimination. A 10-fold variation in brightness of the environment, which happens commonly because of relative cloud cover, does little to influence behavior, but a 10-fold variation in the size of a nearby primate demands immediate attention (Fig. 1.22).

Fig. 1.22 Two separate types of population density maps. A, This is a B-mode map, showing population density as a manifestation of brightness. The problems discussed in the text are readily apparent in this image because it is difficult to ascertain which of the following four cities has the greatest population density: Atlanta, Chicago, Minneapolis, or Dallas. B, This is a bump map showing the same data displayed in terms of height. Here it is possible to easily distinguish population density levels in different cities and to clearly see that New York has the greatest population density by far of all the cities represented. The human eye is simply better designed to analyze height ratios than brightness ratios.

Different ultrasound instruments use different algorithms to relate display brightness to echogenicity, and these are not standardized across instruments. Only in A-mode instruments used for ophthalmologic examinations has there been an attempt by manufacturers to adhere to an imaging standard in this regard. Of course, brightness is also easily manipulated by angle of insonation (see Tissue Anisotropy, as follows), gain or power settings, time gain compensation settings, and selection of transducer frequencies. Further complicating the display of brightness is the influence that tissue parameters have on brightness display; thicker layers of fat, skin, or overlying tissues by necessity will alter the intensity of echoes returning from a deep structure, and it is difficult to compensate accurately for such effects. Given the multiplicity of these variables, the display of echogenicity has been used to primarily enhance resolution rather than as a measure of the intensity of reflected echoes.

Of interest, the amplitude/intensity problem has a parallel in electrodiagnosis. The changes in the amplitude of electrically recorded signals from nerve or muscle tend to convey much less information than do changes in their temporal profile. For example, the placement of surface or needle electrodes has profound effects on amplitude.⁴ Slight movements of surface electrodes may change the recorded compound muscle action potentials by 50% or more, and slight EMG needle movements lead to even more profound motor unit action potential changes. Such movements have relatively little effect on latencies, conduction velocities, durations or interpotential intervals,⁴ parameters, free of such random variation, that are more likely to convey meaningful information regarding pathology than signal amplitudes. Furthermore, selection of different filter settings, thickness of intervening skin and fat, or even directional changes in needle electrodes, all have profound influences on recorded amplitudes in a fashion not dissimilar to that seen with ultrasound. Electrodiagnostic instruments are designed to measure amplitudes (intensity) more precisely than ultrasound instruments, but in fact, even when amplitude is measured accurately, it is not as informative as temporal data.

The brightness problem in ultrasound is a choice of instrument design. It is possible to engineer instruments to reliably measure and record variables such as echogenicity in a standardized fashion, but until there is adequate evidence to suggest that this would be of clinical significance, it will not be pursued. The development of color Doppler and blood flow analysis was a more complicated task. The brightness problem of ultrasound is an accepted limitation of current technology, but the fact that it is rarely a subject of debate or discussion should not imply that future research and improvisation in the area will be unfruitful.

Artifacts

Ultrasound imaging is associated with a variety of artifacts, many of which are self evident and few of which lead to significant problems when interpreting images. It is helpful to identify and name some of the more common artifacts to facilitate operation of the instrument and provide accurate descriptions of image display.

Tissue Anisotropy

Tissue anisotropy can create an apparent artifact, but actually, it turns out to be a rather valuable aspect of ultrasound imaging. Each tissue is different in terms of how it reflects light. Some create much more backscatter ³ (Fig. 1.23; also see Fig. 1.15) and some act more like mirrors, reflecting almost all the incident sound at an angle equal to the angle of incidence in reference to a central axis. Tissues that have high levels of backscattering have low anisotropy and tend to look the same regardless of the angle of incident sound, whereas tissues that act more like a pure reflector and have high anisotropy look much brighter when the transducer is perpendicular and much darker when it is not. Unlike routine MR or CT images in which the directional source of imaging energy is relatively constant, with ultrasound subtle changes in the angle of incidence of the transducer reveal variations in anisotropy in tissue (see Fig. 3.4), can help distinguish different types of adjacent tissues, such as tendon (high anisotropy) and nerve (low anisotropy). This relationship is demonstrated in Figure 1.23. Skilled ultrasonographers use these shifts in transducer angle to provide valuable eye-hand feedback that not only helps interrogate tissue but also interprets what is reflected back.

Fig. 1.23 Cross-sectional images of the median nerve at the wrist. A, An image in which the angle of insonation of the probe is just slightly off the vertical. In this situation the tendons are hypoechoic compared with the nerve. On the right, the transducer is perpendicular to the tendons at the wrist; here they are hyperechoic compared with the median nerve. B, An M-mode trace (shown in the top image, divided by a thin green line) as the transducer is moved from the off perpendicular plane (top figure on the left), through the perpendicular (top figure on right) and off perpendicular in the other direction, and then reversed and brought back to the perpendicular three times. Note that the nerve does demonstrate some anisotropy in that it is slightly brighter when the transducer is perpendicular, but the difference with the tendons is far more striking. The tendency for the human eye to enhance contrast makes it more difficult to see the nerve anisotropy in the B-mode image (A) as opposed to the M-mode image (B).

Shadowing

Other artifacts result from signal attenuation. A classic example occurs with calcified tissues, such as calcific arteries. These tissues do not transmit any sound and either reflect it back to the transducer or backscatter it in other directions. As such, no returning echoes are seen deep to the calcific tissue, but instead a shadow appears with linear edges.³ Shadows are useful because few changes in tissue are dense enough to cause them to appear, and they immediately draw attention to pathology (see Fig. 1.16).

Enhancement

Enhancement is the inverse of shadowing. Some structures, for example cysts, generate no echoes and transmit virtually all the sound energy that is directed through them (see Table 1.2). This tends to make the structures immediately deep to the cyst appear brighter.³ This is because the instrument is programmed to expect a certain loss of sound energy as it passes through tissue, and when this does not occur, the default mechanism of the instrument displays the result as distal enhancement. Recognition of this phenomenon is useful in drawing conclusions about the tissue superficial to the enhancement. Increased echogenicity deep to a structure raises the possibility that it is a cyst or vascular structure (see Fig. 1.17). Color flow Doppler is often helpful in distinguishing between static fluid in a cyst and slow flowing blood, as in the case of arteriovenous malformations or pseudoaneurysms. This distinction is best made before performing a needle aspirate of a suspected cystic lesion.

Other Artifacts

Reverberation artifact occurs when the transducer or some superficial tissue structure acts as a sound reflector (Fig. 1.24).³ The effect, like looking into a three-way mirror, causes multiple equally spaced images of a reflection to appear. These have a naturally artifactual appearance and are generally easy to recognize. Other artifacts occur but generally are self evident. Like their counterpart of optical illusions studied by psychologists, understanding ultrasound artifacts can enhance practical appreciation of the mechanics of ultrasound display.

Fig. 1.24 This is a cross-sectional image of the radius taken at the level of the mid-forearm. On the left, the image is obtained with an 18-MHz probe, and on the right a 10-MHz probe. Note the series of small linear reflections beneath the bone edge. These represent serial reflections of the sound off the transducer and then the bone. More lines more closely interspaced are seen on the left because of the higher frequency of the transducer and the multiple superficial reflectors. The artifact is sensitive to the angle of the transducer because if it is held just off the perpendicular to the bone the sound echoes reflect away from the transducer.

Practical Considerations Regarding Neuromuscular Ultrasound

Selecting an Instrument for Purchase

Almost everyone who uses, or wants to use, ultrasound develops an interest in strengths and weaknesses of commercially available instruments. Because these range in price from $20,000 to $200,000, such interest seems warranted. This portion of the chapter highlights the different commercially available features of an instrument that may influence purchasing decisions.

Cost

Two factors tend to drive the cost of an ultrasound instrument. The first is image quality, the second is the range of functions. The most expensive instruments tend to have both high-quality image displays and numerous ranges of functions. In neuromuscular ultrasound the image quality of superficial soft tissue structures is probably the most valuable feature. Color Doppler blood flow capabilities are secondary, and yet likely to be an important capability as illustrated in subsequent chapters. Range of frequencies, 3D imaging, large display screens, ergonomic utility, gel warmers, multiple transducers, postprocessing capabilities, and portability are of less concern, but can be important to the needs of the individual user. It is sometimes difficult for novice ultrasonographers to fully appreciate the exact needs they have for instrumentation, so leasing a new machine or purchasing a used one may provide useful experience with the technology that can help guide a subsequent purchase. For accounting purposes, most instruments can be assumed to have at least 4 years of effective use, yet many instruments continue to be of value for several more years. Ultrasound technology is advancing at a moderate pace, and investing in the best appropriate technology is critical. Performing an ultrasound examination takes time, and better equipment enhances efficiency, accuracy, and patient care, factors that need to be given due consideration when deciding on how to invest.

Instrument manufacturers in ultrasound vary from large multinational corporations (General Electric) to moderate-sized companies (Esaote) and any number of smaller companies. The largest companies compete in the manufacturing of multiuse premium instruments with multiple features, whereas smaller companies tend to focus on niche markets. Here the competition is less intense, and innovations are more likely to be forthcoming. Larger companies sometimes follow the fortunes of the smaller companies that are competing in niche markets, and buy up the ultimate winner. Product development is continual in newer areas of ultrasound, so careful attention to newer manufacturers, and familiarity with instrumentation principles are of importance in purchasing an instrument for neuromuscular ultrasound.

Service and Training

For novice users in particular, but for all users, access to service and the stability and quality of ultrasound representatives are considerations in selecting an instrument. For physicians in large group practices or academic centers, there is often an ultrasound manufacturer that already has an institutional presence, and this is a good place to start when looking at instruments. An established representative tends to be available and in a strong position to discount the instrument or to provide attractive lease or used equipment options. The representative also provides a standard for comparing other potential vendors (and representatives). An unsatisfactory representative can lead to difficulties that negate many technical advantages available on an instrument, so this aspect of equipment purchase should not be overlooked. It should be noted that relatively few representatives at this time understand the nature of neuromuscular ultrasound, and educating them will help them demonstrate aspects of the instrument that are most relevant to neuromuscular patients.

Instrument Size

Instrument size is problematic because larger instruments use up more space, generate more heat, and are hard to use outside of a central location. If the ultrasonographer intends to share the instrument with another user, or to use it off-site, size is even more problematic. For practical reasons, physicians who wish to use ultrasound and who also perform EMG and NCS will want to have ready access to the instrument in the EMG laboratory. Particularly for new users, this ready availability is critical because it is difficult to resist the temptation of settling for the limitations of electrodiagnosis if it means waiting for access to imaging. Of particular help is the use of a portable battery strip, so the instrument need not be unplugged and restarted as it moves from room to room. Anything that delays mastering the initial learning curve for ultrasound will delay the return on the investment of time and resources in the instrument, so making the instrument accessible is critical. Larger or brighter display screens on some instruments are occasionally of interest, particularly if the instrument is used for teaching residents or fellows. Image quality and resolution are more important than screen size, and many modern instruments are equipped with ports for an LCD projector that can provide broadcasting of images for use in teaching conferences. Ergonomic features are easier to design on the large instrument platforms and are of importance for repetitive use.

Training EMG Technologists in Ultrasound

A trained, nonphysician, ultrasonographer is unlikely to be able to perform neuromuscular ultrasound at first because this is rarely an area of emphasis for most training programs or radiology practices. On the other hand, they bring useful instrumentation skills to the job. Because ultrasonography technologists are among the highest paid technologists in clinical medicine, however, such assistance comes with a price. In the author’s experience, electrodiagnostic technologists can be trained to perform ultrasound of common nerve entrapments, and their ability to correlate electrodiagnostic findings with the results of the electrodiagnostic studies, along with their level of comfort with instrumentation and neuromuscular patients make them quick learners. Ultrasound also helps make them better in electrodiagnosis because it provides feedback on the anatomy of structures that are routinely under study. As with any procedure, however, it is tantamount that the interpreting physician have sufficient experience and training to be able to perform the procedures independently, supervise technician conduct of procedures, troubleshoot technical problems, keep up with advances in instrumentation, and apply new discoveries as they evolve.

The American Association of Neuromuscular and Electrodiagnostic Medicine (AANEM) is working on standards of training suitable for physicians who wish to perform neuromuscular ultrasound.⁴²

Safety

Ultrasound is the safest imaging modality in use, and there are few, if any, reports of any physical harm resulting from the technique.³ It is the standard imaging procedure for looking at the fetus. The theoretic risk of ultrasound relates to its ability to heat tissues. Of tissue, perhaps the most vulnerable to ultrasound is the eye.⁴³^–⁴⁵ With even very brief bursts of high-intensity ultrasound, cataracts may be induced. Imaging the optic nerve or eye requires that the ultrasonographer be well versed in the proper instrumentation for the procedure and its risks. Temperature probes or hydrophones are available to measure heat increases in tissue and ambient sound energy for those wishing to quantify thermal effects of ultrasound imaging.