Physics of Ultrasound





KEY POINTS





  • Ultrasonic waves are elastic, mechanical waves.



  • Ultrasound imaging is safe and noninvasive.



  • Imaging provides real-time visualization.



  • Contrast imaging improve vascular detection.





Characteristics of Ultrasound Imaging


Ultrasonic waves are elastic, mechanical waves. Ultrasound requires an elastic or viscoelastic medium for propagation. In an element of an elastic medium, longitudinal stress (i.e., force per unit area) applied to the medium is proportional to the strain (i.e., change in the length divided by the initial length). This relationship is described by Hooke’s law. In a viscoelastic medium, stress is related to strain amplitude and to the rate of strain variation.


There is mechanical interaction between the wave and the body. The wave progressively decreases in amplitude, or is attenuated , as it propagates through the medium because of energy lost through scattering from the principal propagation path and lost as heat. Attenuation of the wave increases with increasing frequency. The higher the ultrasound frequency, the more limited is the accessible imaging depth.


Ultrasound imaging is safe and noninvasive. As the ultrasonic pulse passes through a medium, molecules within the body are reversibly perturbed from their equilibrium positions. Images are formed by sensitively detecting echoes of the pulse returned from interfaces or from scattering structures within the tissue. The relative intensity of a returned echo is represented in terms of relative pixel brightness. The sound path’s orientation and the echo time-of-flight are used to map each echo intensity to a position in a reconstructed image of the medium.


Systems are relatively low cost. Ultrasonic transducers used to initiate and to detect ultrasonic pulses, and the electronics that accompany them are relatively low cost compared with the equipment necessary for most other imaging techniques. This is not a direct result of physics, but physics has helped indirectly by providing an elegant solution for the production of ultrasonic waves (i.e., piezoelectric crystals) within the frequency and amplitude ranges required for medical imaging.


Imaging provides real-time visualization. Ultrasonic probes can be constructed from a tightly spaced row of small piezoelectric elements that generate and detect ultrasound energy (i.e., transducer array). Transmission and reception channels are switched on and off sequentially from one group of elements to another to interrogate along different sound paths. Echo-based imaging speed is limited by the pulse’s travel time to and back from the deepest part of the imaged zone (on the order of 0.1 ms for an 8-cm depth). Movements within structures can be observed as they occur. The velocity and direction of blood flow or other tissue displacements can be evaluated.


Specific imaging modes are applied for contrast agent detection. Ultrasound contrast imaging for better delineation of vascularized zones relies on the intravenous injection of stabilized, micrometric gas bubbles. Because microbubbles can oscillate in a unique, nonlinear way in response to an ultrasonic pulse, they can be detected while virtually ignoring the background signal from surrounding tissues. Contrast-specific detection allows visualization of microvascular perfusion that cannot be observed by other ultrasonic techniques.


Numerous applications take advantage of the noninvasive and real-time nature of ultrasound. It is used by many specialists, such as obstetricians, cardiologists, radiologist, gastroenterologists, endocrinologists, and surgeons. Rheumatologists apply ultrasound predominately for musculoskeletal evaluation.




Ultrasound Mechanisms and Settings


Pulses and Echoes


Imagine a very crowded train platform during a major transportation strike. At the edge of the crowd, somebody shoves. Each person is rooted to the spot he or she occupies by the force of the surrounding crowd. Nevertheless, the shove travels through the crowd. When it gets to me, I move a little bit forward (shoving my neighbor), pull myself back the other way, and eventually return to my original position. At this point I am possibly a little heated with irritation—useless energy loss—and probably wondering how I can reach my cell phone with my hands pinned to my sides.


If the crowd were homogeneously composed of an infinite number of average-sized, mild-mannered people like myself, the shove would probably continue traveling a long time and across a large distance. But it is not. Eventually, the shove encounters something different from the rest, such as someone who is a little fed up. When this happens, at least part of the shove may be returned along its original path. As it passes me again, I am perturbed in the opposite direction from last time (likely dropping the cell phone I retrieved from my pocket). The returned part of the shove continues back toward its source.


Speed of Sound and Depth Range


In the loose analogy presented, each person represents a molecule in a fluid-like medium. The pulse is transmitted by a sequence of impacts between neighboring molecules put into brief oscillation by the pulse, but the individual molecules return to their original equilibrium positions after the passage of the pulse. Transmission of the pulse from one point in the medium to another takes some finite time, and the speed of this transmission is referred to as the speed of sound .


When the transducer is placed against the body, the round-trip time needed for the pulse to travel from the transducer to the depth of some structure and for the structure’s echo to return to the transducer is described as follows:


round-trip time = 2 × depth to structure average speed of sound


Because of the very modest variations in the speed of sound (only a few percent) between different types of soft tissues ( Table 1-1 ), pulse-echo depth-ranging can be performed by assuming the average speed of sound to be 1540 m/s ( Fig. 1-1 ).



Table 1-1

Wave Propagation Speed and Acoustic Impedances for Materials Encountered or Used in Ultrasonic Imaging




























































Medium Speed of Sound (m/s) Acoustic Impedance (megaRayls MKS )
Materials encountered when imaging the body
Muscle 1580 1.645-1.7
Liver 1570-1578 1.65-1.66
Kidney 1560 1.62-1.64
Vitreous humor (eye) 1520 1.52
Blood 1550-1584 1.61-1.68
Water at 20° C 1482 1.48
Fat 1430-1450 1.33-1.38
Bone 3198-3500 6.36-7.8
Air 330 0.0004
Piezoelectric materials used in transducer probes (sample values)
Quartz 5000 13.3
Polymers (PVDF) 2200 3.92
Composites (1-3) 3000 18.0

PVDF, polyvinylidene fluoride; Rayls MKS , kg m –2 s –1 .

Data from Webb A: Ultrasound imaging. In Webb A (ed): Introduction to Biomedical Imaging. Hoboken, NJ, John Wiley & Sons, 2003:107-156; Szabo TL (ed): Diagnostic Ultrasound Imaging: Inside Out. London, Elsevier Academic Press, 2004.



F igure 1-1


D epth range .

The ultrasound pulse travels to the depth (D) of a structure, and the echo returns in a time interval of Δt. The structure’s depth can be estimated by multiplying Δt/2 by the average speed of sound (approximately 1540 m/s in soft biologic tissue).


Acoustic Impedance, Reflection, Transmission, and Refraction


The presence of the angry person in the otherwise homogeneous crowd represented a change in the propagation medium’s acoustic impedance (i.e., material density multiplied by the speed of sound propagation). For perpendicular incidence of a wave at an infinite, smooth interface, the fraction of the incident amplitude or intensity that will be reflected and the fraction that will be transmitted depend on how much the acoustic impedance varies between the materials on either side of the interface ( Fig. 1-2 ). If the incidence is not perpendicular but at an angle with respect to the normal axis of the interface, the angular direction of the transmitted beam will be modified at the interface, or refracted . Effects of refraction on the propagation path usually can be neglected in medical imaging, except for nonperpendicular propagation across interfaces between materials with strong differences in the speed of sound, such as can occur at a soft-tissue-to-bone interface or for certain interfaces in the eye.




F igure 1-2


T ransmission and reflection .

Consider a smooth boundary between two types of tissue. The boundary extends out laterally to a distance much greater than the ultrasound wavelength. The acoustic impedance (Z) of each tissue is equal to the density (ν) times the speed of sound ( c ). A, A wave with intensity I i is perpendicularly incident on the boundary. B, The reflected and transmitted intensity is determined by the acoustic impedance values of the two tissues. The greater the difference in impedance, the stronger the reflected intensity.


The approximate acoustic impedance for some of the materials that can be found in the human body is given in Table 1-1 . A few practical consequences can be inferred from comparing these acoustic impedance values. There is a significant difference between the acoustic impedances of bone or air and other materials in the body. Any bone or air-containing structures along the sound path strongly reflect the ultrasonic pulse ( Fig. 1-3 ) and keep all or most of it from penetrating to deeper structures. The line of sight for imaging needs to be chosen to get around any such structures between the surface of the body and the imaged zone. Reflections from interfaces between different types of soft tissue usually are less than 0.1% of the incident intensity, providing weak but detectable echoes while leaving considerable energy in the wave for propagation to deeper zones.




F igure 1-3


A rtifacts from strong reflectors .

Dorsal forearm views include a short-axis cross section of the radial bone. A, When ultrasound is strongly reflected at the tissue-to-bone interface, much less intensity is transmitted to deeper regions. This is called acoustic shadowing (S). B, Very strong echoes from the bone surface can reverberate (R) between tissue interfaces (arrows) or between the transducer face and the strong reflector. This produces a bright echo trail posterior to the tissue-bone interface. The echoes occur at regularly spaced intervals related to the distance between the reflective interfaces that give rise to the reverberations.


Piezoelectric materials used in ultrasound probes typically have very different acoustic impedance from that of body tissue (see Table 1-1 ). To improve efficiency of transmission from the ultrasound probe to the body, matching layers are hidden in the probe’s tip. Matching layers are made of materials with intermediate acoustic impedance and are the right thickness to minimize parasite signals from internal reflections. Ultrasound coupling gel is also placed between the patient and the transducer to provide an air-free, fluid-like joint, across which the operator can move and align the transducer. By arranging the system for efficient transfer of ultrasound from the source to the medium, the acoustic coupling and the image are improved.


Piezoelectric Materials and Their Resonant Frequencies


The ultrasonic probe is called a transducer , and it uses some interesting material physics discovered in 1880 by Pierre and Jacques Curie. Certain materials (e.g., quartz crystals, ferroelectric crystals, piezoelectric ceramics, piezoelectric composites, piezoelectric polymers) have an atomic structure that is organized such that when the material is deformed mechanically (along a specific orientation), it takes on a net charge (i.e., direct piezoelectric effect). A distortion of the material can be detected by placing electrodes on its opposite faces and measuring the voltage potential (i.e., different voltage levels on the two electrodes). If a voltage potential is applied across the electrodes, the material will deform to adjust its internal electric field to the new situation (i.e., reverse piezoelectric effect). A probe constructed of this type of material can be used to produce and to detect vibrations.


If a difference of potential is applied with a cyclic and continuous variation of the positive and negative poles, the piezoelectric material will undergo successive contraction and expansion. If the frequency of this cyclic voltage variation is equal to the resonant frequency of the piezoelectric material, the mechanical vibration produced by the varying voltage potential is optimized and has much greater amplitude than at nonresonant frequencies. The resonant frequency of a piezoelectric wafer is a function of its thickness and the speed of sound in the wafer.


Frequency, Wavelength, Damping, and Bandwidth


Electrically “tapping” a piezoelectric wafer is a bit like tapping a fine crystal glass. It does not respond with a vibration as short as the tap that set it in motion. Instead, it rings for some time at a particular frequency determined by its geometry. If a mechanically isolated piezoelectric wafer is tapped with a rapid variation of the voltage across its faces, it will vibrate at a frequency centered on its resonant frequency and continue vibrating for some time. Ringing of a glass can be stopped more quickly by placing a finger on its edge to dampen the vibration, and much of the energy that was transferred by the original tap is lost. There is a trade-off between maximizing energy transfer into the medium (needed to sensitively detect structures in the body) and the briefness of the ultrasonic vibration (needed to make a pulse short enough to provide good spatial localization of structures). Medical imaging transducers typically produce pulses damped to between 1 and 3 cycles of vibration that contain a broad gaussian distribution of frequencies, typically centered on the resonant frequency of the piezoelectric wafer. Doppler pulses are typically less damped and can be 5 to 20 cycles long.


The oscillation of the crystal modifies the pressure at the surface of an acoustically coupled medium ( Fig. 1-4 ) by cyclically pushing and releasing. As the longitudinal wave propagates in the medium, zones where the molecules are pushed slightly closer to each other than usual (i.e., zone of compression ) and zones where molecules are a bit stretched out compared with the equilibrium spacing (i.e., zone of rarefaction ) advance farther and farther into the medium. The parts of the wave that act on the medium in the same manner (i.e., same direction and amplitude of pressure modification) have the same phase (see Fig. 1-4 ). The distance between two points of a wave with the same phase is the wavelength (λ) in the medium. The linear frequency (ν) and the wavelength in the medium are related by the speed of sound propagation in the medium.




F igure 1-4


P ressure wave in a medium .

Vibrations (ν cycles of vibration per second) from the ultrasonic transducer are transmitted into the medium. Zones of material compression and rarefaction result and advance by neighbor-to-neighbor vibrations. The distance between two zones of like phase (peak rarefaction has been selected in this diagram) determines the ultrasonic wavelength (λ) in the material. The λ equals the speed of sound in the material divided by the frequency of transducer vibration (ν).


Medical imaging transducers vibrate at resonant frequencies in a range from about 1 to nearly 20 MHz (1 to 20 million vibrations per second); the linear frequency is the number of vibrations per microsecond (1 MHz = 1 vibratory cycle per μs). This frequency range is a small part of the total range of frequencies defined in acoustics as ultrasound, which spans from about 20 kHz (20,000 vibrations per second) to 10 13 Hz (thermal vibration frequency of molecules). This range is deemed ultra based on the fact that sound vibrations at frequencies above 20 kHz cannot be detected by human ears. To understand the consequences of choosing the 1- to 20-MHz frequency range for ultrasound medical imaging, the relationships among frequency, spatial resolution, and imaging depth must be considered.


Spatial Resolution


The spatial distribution of the acoustic energy (i.e., beam dimensions) varies as a function of distance from the acoustic source and along each of the three orthogonal axes of the image ( Fig. 1-5 ). These beam dimensions at the transducer focus determine the minimum spatial separation between two structures necessary to discriminate, or resolve, them from each other. The size of structures with dimensions smaller than or on the order of the resolution cell cannot be well evaluated.




F igure 1-5


R esolution cell .

A typical resolution cell is diagramed for the focal zone of a linear array transducer. The best spatial resolution is along the axis of propagation (i.e., axial resolution). Good lateral resolution can be obtained with time-delay focusing techniques. Resolution usually is most limited in the dimension described by the image slice thickness (i.e., elevational resolution).


Axial Resolution


Because the frequency is related to the wavelength in the material and because the pulse is damped to n λ wavelengths (n λ is not necessarily an integer), the frequency choice is central to the attainable axial resolution ( Fig. 1-6 ). Structures separated along the axis by a distance greater than or equal to the axial resolution limit will give rise to well-separated echoes that should appear at distinct and visually separable depths in the reconstructed image of the medium. For a 1.5-wavelength pulse at 1-MHz center frequency propagating in soft tissue with a speed of sound of 1.54 mm/μs, the axial resolution is on the order of 1 mm. For a similar 10-MHz pulse, the axial resolution is on the order of 0.1 mm.




F igure 1-6


A xial resolution .

Consider two structures at different distances from the transducer along the pulse propagation path. A, Echoes from two structures separated by a distance (D) greater than the axial resolution (pulse length/2) can be discriminated. B, Echoes from two structures separated by a distance smaller than the axial resolution cannot be discriminated. Pulse length is equal to the number of cycles per pulse (n λ ) times the wavelength (λ). Pulse length can also be expressed as the number of cycles per pulse times the speed of sound ( c ) divided by the pulse frequency (ν).


Lateral Resolution


The minimum separation necessary to discriminate structures that are side by side and lying in a plane perpendicular to the axis of wave propagation and within the imaging plane depends on the wave phenomenon of diffraction . The ultrasonic transducer or transducer array has some finite size and shape. The pressure amplitude at each point in the acoustic field results from the sum of the amplitudes of all the wavelets generated at different points along the transducer surface that arrive at the same spot at the same time. The sum is higher at points in space where the wavelets arrive in phase with each other and lower at points where the wavelets arrive out of phase. Along the central axis of wave propagation, there is a final point with maximum constructive interference. At this distance from the transducer, the beam is effectively strongest and most focused. Moving away from the central beam axis to one side, the interference is less constructive, and intensity progressively falls off; moving out from the central axis, after passing through a minimum intensity, there is a much smaller local maximum intensity associated with an off-axis zone of constructive interference, and this gives rise to a lateral lobe. The lateral resolution is related to the distance from the central axis where the intensity or amplitude of the returned echoes is within 6 dB of the peak on-axis value. A 6-dB decrease is half-peak amplitude or one fourth of peak intensity: –6 dB = 20 × log 10 [(0.5 × A max )/A max ] = 10 × log 10 [(0.25 × Imax)/Imax]. In practice, lateral resolution of linear array ultrasound imaging systems can be evaluated using commercially available scattering phantoms containing fine wires with calibrated lateral spacing.


Elevational Resolution


The intensity of the beam also falls off progressively as a function of the distance from the central plane of the imaged slice of tissue. In other words, the actual slice of tissue giving rise to echoes has a certain thickness to it. In general, linear arrays are cylindrically focused to minimize this thickness near the focal zone. The elevational resolution is related to the width of the beam along the dimension of the slice thickness with echo intensity or amplitude within at least 6 dB of the peak value.


Focusing


The focal distance can be modified by changing the geometry of the transducer’s surface (e.g., curving it) or by applying relative time delays to excitation of individual array elements ( Fig. 1-7 ). This changes the relative time delays between arrival of the wavelets from different parts of the source at each region of the field, and at the point where the pressure amplitude is maximum, it is shifted along the axis. For simple spherical focusing, the 6-dB lateral resolution limit is as follows:


Lateral resolution limit = c ν × d focal 2 · a

Only gold members can continue reading. Log In or Register to continue

Stay updated, free articles. Join our Telegram channel

Mar 1, 2019 | Posted by in RHEUMATOLOGY | Comments Off on Physics of Ultrasound

Full access? Get Clinical Tree

Get Clinical Tree app for offline access