Fundamentals of Neuromuscular Ultrasound

Neuromuscular Ultrasound History

The advent of medical ultrasound dates to the 1940s when the first experimental devices were used to image the brain through the skull to assess for tumors, and the abdomen to assess for gallstones. Over the years, ultrasound has become widely used in obstetrics, as well as in the evaluation of abdominal structures (especially liver, kidney, and gallbladder). The use of ultrasound has also become common place in cardiology, with the everyday use of transthoracic and transesophageal echocardiograms. More recently, ultrasound is being increasingly used by anesthesiologists to localize nerves for regional anesthesia and guide the injection of anesthetics. Ultrasound is also now routinely used in emergency departments, especially in cases of trauma. In musculoskeletal (MSK) disorders, ultrasound is used to assess tendons, ligaments, joints, and bone and is often used to look at the shoulder joint and rotator cuff. Although there is a fair amount of overlap between the structures seen on MSK and neuromuscular ultrasound, they focus on different things. Neuromuscular ultrasound focuses specifically on peripheral nerves and muscles and their surrounding structures.

The use of ultrasound to study neuromuscular conditions was first introduced by Heckmatt, Dubowitz, and Leeman in 1980 when they studied young boys with Duchenne muscular dystrophy. They noted profound changes in the muscle of patients compared with normal controls ( Fig. 17.1 ). By today’s standards, the technology and resolution of these early images were primitive. Over the ensuing years, ultrasound was occasionally used by a small number of neuromuscular physicians in selecting optimal sites for muscle biopsy but was not widely adopted. However, over the past 10–15 years, the use of neuromuscular ultrasound has expanded widely to become a validated and important diagnostic tool. Many recent advances account for this more widespread use of neuromuscular ultrasound. The software has markedly improved, as have the ultrasound probes, with the ability to use much higher frequencies resulting in a much higher resolution of images. The size of the machines has dramatically reduced, making them very portable ( Fig. 17.2 ). Thus, one ultrasound machine can easily be shared among several laboratory rooms. In addition, the machine can easily be taken to inpatient floors. The cost of ultrasound machines has come down greatly over the years and is now comparable to most electromyography (EMG) machines. Indeed, some manufacturers are working on devices that house both EMG and ultrasound machines in one unit.

In addition, neuromuscular ultrasound has many other clear advantages. It is painless. It is safe, with no side effects. It is dynamic: one can visualize nerves, muscles, and tendons as a limb is moved either actively or passively to help understand the relationship between the nerves and muscles examined and their surrounding structures. Research in neuromuscular ultrasound has greatly expanded with several thousand peer-reviewed articles published each year. Neuromuscular ultrasound workshops are offered by several professional societies and universities. There are now several textbooks dedicated specifically to neuromuscular ultrasound. This text is not meant to be a substitute for a comprehensive textbook on neuromuscular ultrasound. Rather, in this textbook, we highlight the basics of neuromuscular ultrasound and many of its most useful features as a complement to EDX testing, especially in some of the later clinical chapters. We refer the reader to several excellent textbooks on ultrasound and neuromuscular disorders in the suggested readings.

It is important to emphasize that neuromuscular ultrasound does not and will not replace EDX testing but rather is complementary to it. EDX testing assesses the physiology of nerve and muscle, which ultrasound cannot do, and often localizes the problem. In contrast, ultrasound is an imaging test that can also often localize the problem but can also add some specific diagnostic information that EDX studies cannot detect. For instance, if one takes a history from a patient who has pain and paresthesias in their fingers that awakens them from sleep, one might quickly come to the likely clinical diagnosis of carpal tunnel syndrome (CTS). In the EDX laboratory, median neuropathy across the wrist can usually be demonstrated easily on nerve conduction studies (NCSs), including the pathophysiology (demyelination vs. axon loss). However, even though the clinician can diagnose CTS clinically and demonstrate median neuropathy across the wrist on EDX studies, the question remains: what is causing the median neuropathy? Is it simply wear and tear and enlargement of the transverse carpal ligament? Or does the patient have tenosynovitis? A ganglion cyst? A nerve sheath tumor? Or deposition of abnormal tissue (e.g., amyloid or synovial hypertrophy) adjacent to the carpal tunnel? And the differential diagnosis of potential structural causes does not end there. This is where ultrasound can add information that EDX studies cannot assess. Again, it is the neuromuscular physician who is the optimal person to perform the ultrasound study in conjunction with the EDX study. Once the information from the EDX study is known, ultrasound can then be used in a directed manner to obtain important additional structural and dynamic information.

Basic Physics of Ultrasound

Ultrasound literally means “beyond sound.” Everyone is familiar with sound. Sound is defined as a mechanical energy that is transmitted by longitudinal pressure waves within a medium. Sound waves are measured in Hertz (Hz). One Hertz is one cycle per second. Most humans can hear sound with frequencies between 10 Hz and 20,000 Hz. Humans cannot hear sound above 20,000 Hz, although dogs can hear sounds up to 40,000 Hz; hence the “dog whistle”—a whistle that creates a sound wave between 20,000 Hz and 40,000 Hz that can be heard by a dog but not a human. In ultrasound, the frequencies are much, much higher than a dog whistle. They are not in the thousands but rather in the millions of Hz. One million cycles per second is a megahertz, abbreviated as MHz. Most neuromuscular ultrasound probes have frequencies between 2 MHz and 20 MHz. In addition, there are some ultra-high frequency ultrasound devices now in use that operate in the 50- to 70-MHz range.

Ultrasound machines depend on the property of piezoelectricity. Piezoelectric materials, which are contained in the ultrasound probe, create sound waves when a voltage is applied to them. Thus, they transform electrical energy into mechanical (sound) energy. However, they also do the opposite: when sound energy is absorbed into piezoelectric material, that energy is transformed into electrical energy. Thus, in medical ultrasound machines, electricity is applied to piezoelectric elements in the probe, which create sound waves that travel through body tissues. Sound waves travel at different speeds through different media. In air, sound travels at approximately 330 m/s. When one knows how fast sound waves travel through the air, the distance can be calculated between the source generator of the sound and where the sound is detected, using the simple formula Rate × Time = Distance. For example, when you see a bolt of lightning and then hear the thunder 5 seconds later, you can use this simple formula to calculate how far away the lightening is from you. By knowing that it took 5 seconds for the thunder to travel to you and that sound travels through air at 330 m/s, then the distance can be calculated as 330 m/s × 5 seconds = 1650 meters. One knows that the lightening was approximately 1 mile away as 1 mile is 1609 meters.

In most body tissues, sound travels at approximately 1540 m/s. When an ultrasound wave travels through tissue, it can do different things ( Fig. 17.3 ). In some cases, it simply continues through that tissue and is eventually absorbed as heat. Indeed, there are therapeutic (as opposed to diagnostic) ultrasound devices that are purposely used to create heat deep in tissues, for therapeutic purposes. Every tissue has an “acoustic impedance.” If the sound wave encounters an area where two tissues with different acoustic impedances lie adjacent to each other, echoes are created. These echoes are then transmitted back to the piezoelectric elements in the probe to be converted back to electrical energy. Thus, echoes are created when sound waves encounter a boundary of two dissimilar tissues. The strength of the echo increases as the difference between the two acoustic impedances increases. Accordingly, if there is a marked difference in impedance between two tissues, very strong echoes occur. Conversely, if there is little difference between two adjacent tissues, very weak or no echoes occur. If the echoes are directed back toward the probe, they will reach the probe and can be recorded. These echoes are what create lines and images on the ultrasound. By knowing the sound wave’s speed of travel in tissue (1540 m/s) and measuring the time an echo is received back after an ultrasound pulse is given, one can then calculate the distance (i.e., the depth) of the tissue that created the echo. Information from both the location (depth) of the tissue creating the echoes and the intensity (brightness) of the returning echoes helps create the ultrasound image. Because the sound wave must travel to the tissue that created the echo and back, the time at which the echo occurs represents twice the distance to the tissue. Again, this is important as the ultrasound image is created based on the distance traveled and the intensity of the echo.

However, if the tissue is at an angle to the sound wave, it can reflect the echo back in a different direction and may never reach the probe. If the boundary where the echo is created is irregular, echoes can scatter in many different directions. This back scattering results in “speckle” of the ultrasound image, which can interfere with echoes of interest.

On ultrasound, areas that are bright are known as hyperechoic . Those that are dark are hypoechoic . Complete absence of echoes resulting in complete blackness is termed anechoic .

Ultrasound can be used in different modes. If an ultrasound probe had only one piezoelectric element, it would result in a single line of ultrasound information, known as amplitude mode (A-mode) ultrasound ( Fig. 17.4 ). In the amplitude mode, each spike corresponds to an echo. The height of the spike corresponds to the strength of the echo, and the time of the spike correlates to the depth of the tissue that creates the echo. To create an image, ultrasound probes contain hundreds of piezoelectric elements that are arranged in a row. Thus, hundreds of individual lines of ultrasound information are recorded simultaneously, which can be stitched together digitally to create a grayscale image. This is termed brightness mode (B-mode) ultrasonography, which is the most common ultrasound image used (see all ultrasound figures in this chapter after Fig. 17.8 ). In addition to the set of piezoelectric elements, ultrasound probes contain other layers and materials ( Fig. 17.5 ) that couple and focus the ultrasound beam in addition to dampening undue vibrations. The piezoelectric elements act as both sender and receiver. They are continually sending out sound waves and then turning off, waiting to receive echoes back. Indeed, most of an ultrasound cycle is spent with the elements in the receiver mode.

Another mode, motion mode (M-Mode), combines A-mode and B-mode. On M-mode ultrasound, a standard B-mode image is displayed with a line placed on the structure of interest (the index line) ( Fig. 17.6 , top image, green arrow). At the same time as the B-mode image is being displayed, the M-mode image records the ultrasound information from the index line continuously over time ( Fig. 17.6 , bottom image). One can adjust the placement of this index line. M-mode is most useful in looking at tissue movement (as in cardiac function, or in movement of the diaphragm). In Fig. 17.6 , the bright line behind the liver in the B-mode image (top image, red arrow) is the echo created from the diaphragm with which the index line overlaps. In the M-mode trace (bottom image), it is actually A-mode information from the index line (amplitude and depth) that is recorded over time. The bright line in the bottom image is the echo created from the diaphragm, in this case showing its movement over time (bottom image, red arrow). M-mode is most helpful in neuromuscular ultrasound when assessing muscle movement over time. For example, a useful measure of diaphragmatic function is to use M-mode neuromuscular ultrasound to measure diaphragmatic excursion over time during inspiration and expiration.

Probes

There are several types of probes used for ultrasound ( Fig. 17.7 ). The most common include the linear probe, the high-frequency “hockey-stick” probe, and the curvilinear probe. The linear probe is the work horse of neuromuscular ultrasound. It generates a rectangular image with maximal frequencies typically in the 12–16 MHz range ( Fig. 17.8 , left). The smaller so-called hockey-stick probe is also frequently used in neuromuscular and MSK studies. It has a smaller footprint with typically higher frequencies in the range of 18–20 MHz. This probe is used to look at small structures, especially those very close to the surface. Because of its small footprint, it is also very useful around body areas that are uneven or have bony protuberances. Lastly is the curvilinear probe ( Fig. 17.8 , right). This probe generates a sector image (like a piece of pie) as opposed to a rectangular image and is much lower in frequency (typically in the 2–5 MHz range). Curvilinear probes are used when studying very deep structures, such as the sciatic nerve at the gluteal fold and especially the diaphragm behind the liver.

Note that there are fairly significant differences in the frequencies of these different probes. In general, as the frequency goes up, the resolution goes up. Thus, higher-frequency probes are able to discern much smaller structures. However, there are tradeoffs for higher frequencies. The higher the frequency, the greater the attenuation of the image as the wave travels through tissue. Thus, high-frequency probes are only useful for structures that are close to the surface (typically no deeper than 3–4 cm), as there will be little attenuation of the image. At greater depths, one needs to lower the frequency of the probe being used or use a probe with a lower frequency such as the curvilinear probe. Hence, for example, abdominal ultrasound is done with curvilinear probes (which have much lower frequencies), as much greater depths are needed.

Optimizing the Image

When performing an ultrasound study, there are several choices and adjustments to make to optimize the images ( Box 17.1 ). First, the proper probe needs to be selected. For most neuromuscular ultrasound studies, the standard linear probe is used, with a frequency of at least 12 MHz. Ultrasound gel must be used between the probe and the skin. A generous amount of gel should be used to eliminate any air between the probe and the skin. Air has an incredibly high acoustic impedance through which ultrasound waves cannot travel. The probe should be held lightly against the skin, as too much pressure will displace the underlying gel.

To obtain an ultrasound image, the first step is to set the depth. If the depth is set too deep, much of the image will be taken up by black areas below the area of interest where the ultrasound beam has been completely attenuated ( Fig. 17.9 , top image). If the depth is set too shallow, one may unintentionally cut off part of the object of interest and its surrounding structures below ( Fig. 17.9 , bottom image). The second step is to adjust the ultrasound focus ( Fig. 17.10 ). Similar to a camera, the ultrasound machine can focus the sound waves at a particular depth, called the focal depth, to best see images at that depth. The focal depth is adjustable on all ultrasound machines. Indeed, one can set more than one focal depth. However, the downside of setting increased numbers of focal depths is that it decreases the frame rate (the rate at which the image refreshes itself). Slower frame rates result in blurring of the image as the probe is moved. The third major adjustment is the brightness ( Fig. 17.11 ), which is basically an adjustment of gain or sensitivity. The brightness setting does not change the power of the ultrasound waves sent out, but more or less amplifies the returning echoes. The brightness setting should be adjusted so that the image is neither washed out nor too dark such that important details cannot be seen. The fourth adjustment is the frequency. The highest frequency will result in the best resolution. However, the higher the frequency, the more attenuation of the ultrasound beam as it moves deeper. If deeper structures are not well seen, one should either change the probe to a lower-frequency probe or lower the frequency of the probe being used until the deeper structures of interest are well seen. All commercial medical ultrasound machines come with presets for all parameters, which are already optimized depending on which structures are being viewed (e.g., wrist, ankle, etc.). These preset settings are an excellent starting point and often will need little to no adjustment for the most common studies done.

One should next make use of the color Doppler option ( Fig. 17.12 ). From knowledge of basic physics, it is known that the Doppler effect occurs when a source that is producing sound moves toward a receiver, resulting in the sound waves being compressed and the frequency going up. Conversely, as the source that produces sound moves away from the receiver, the frequency goes down. Ultrasound takes advantage of the Doppler effect in the assessment of moving tissues. When a Doppler signal is present (or positive), red indicates that the sound source is moving towards the probe, and blue that it is moving away. Color Doppler is especially useful for looking at blood flow. Blood flow in arteries is usually well seen on Doppler, whereas blood flow in veins is typically not seen unless the amount of flow is substantial. For Doppler ultrasound to work, the probe must be at least partially directed at an angle to the flow of the blood. If the probe is positioned at 90° to the direction of flow, there will be no Doppler effect. The probe must be tilted either toward or away from the vessel to see blood flow. Note, however, this limitation of color Doppler can be overcome by the use of a power Doppler ( Fig. 17.13 ). Power Doppler is very sensitive to the presence of movement of red blood cells, without being direction or angle specific. It is particularly useful for small arteries and those vessels with low-velocity flow. Power Doppler displays a characteristic deep orange color on imaging.

Next, the proper orientation of the images must be maintained. Every ultrasound probe has some type of ridge, button, or symbol on one end of the probe that corresponds to the upper left of the display screen. By convention, when doing a longitudinal study, the left of the screen should always be facing cephalad. When obtaining an axial image, by common convention, the left side of the screen corresponds to the right side of the patient, similar to the convention for magnetic resonance imaging (MRI) and computerized tomography scans (i.e., the anatomic position). Using this approach, the image orientation results in mirror images comparing the left with the right side ( Fig. 17.14 , top image). However, an alternative is to have the left side of the screen always facing the patient’s lateral side. When using this approach, comparing one side to the other, the orientation will remain the same for both left and right ( Fig. 17.14 , bottom image).

One of the major advantages of ultrasound is its ability to look at structures in many different planes. The conventional planes are axial, sagittal, and coronal planes, which are easily visualized on ultrasound ( Fig. 17.15 ). Indeed, when assessing a finding on ultrasound, it is good practice to image that structure in two different planes. This is similar to an MRI scan or the standard posteroranterior (PA) and lateral chest X-ray, which image structures in more than one plane. When an abnormality is visualized in more than one plane, one can be more assured that the findings are real (e.g., think about the cases where a lung nodule appears on a PA chest X-ray but not on the lateral chest X-ray and turns out to actually be a nipple shadow in the skin).

Images can also be designated as being viewed from either the short or long axis of a structure ( Fig. 17.16 ). The short axis is also known as transverse and the long axis as longitudinal. Orientations between transverse and longitudinal are known as oblique. It is important that one denotes what type of image orientation is being recorded when obtaining an ultrasound image. During an ultrasound examination, one should save many images, especially of abnormalities, ideally in different planes. In addition, it is also very helpful to save brief cines (i.e., movie clips) of the examination. Indeed, when looking at a static picture of an ultrasound image, it is sometimes difficult to identify the structure that the image was taken to show. For instance, it is not uncommon that during an examination, such as when following a nerve, the screen is frozen to take a screenshot of the nerve for measurement. However, if you look away from the screen for a moment and then look back, the nerve may no longer be so obvious. Indeed, it is the constant moving of the ultrasound probe that makes many objects become conspicuous. Most ultrasonographers are constantly moving the probe for this very reason.

Common Tissue Patterns

There are several normal tissues that must be recognized on every neuromuscular ultrasound. Each has a specific shape, brightness, and pattern, in addition to other characteristics.

Peripheral Nerve

To understand the ultrasound appearance of peripheral nerve, knowledge of its microscopic anatomy is needed. Peripheral nerve has three layers of connective tissue: the epineurium, perineurium, and endoneurium ( Fig. 2.8 ). Individual axons are separated by endoneurium. Many axons are grouped into bundles known as fascicles that are surrounded by perineurium. Multiple fascicles make up the peripheral nerve that is surrounded by epineurium. When one looks at a transverse image of the nerve on ultrasound, it has a characteristic “honeycomb pattern” ( Fig. 17.17 , top image). The actual nerve fibers are dark (i.e., hypoechoic), surrounded by the connective tissue of the perineurium. Connective tissue, including the perineurium and epineurium, is bright (i.e., hyperechoic) on ultrasound. When the probe is oriented for a longitudinal image, the bright epineurium is seen on the periphery of the nerve (above and below the nerve), with bright parallel lines inside, which represent the perineurium ( Fig. 17.17 , bottom image). When color Doppler is placed over a nerve, typically little or no signal is seen because the blood flow to nerves is in small vessels, which are below the range discernible by color Doppler.