Physics and Knobology

A basic understanding of the physics underpinning ultrasound is crucial for undertaking a formal exam in echo and also (more importantly!) be able to obtain quality images when undertaking a scan. Some of the pitfalls leading to interpretation errors stem from a poor understanding of how optimising depth, frame rate, gain etc… can influence overall quality. All artefacts on the echo image can be explained by the physical principles of ultrasound and so a good understanding is a necessity!

Recent technological advances in acquiring 2D and 3D echo images alongside improvements in processing power and the miniaturisation of probe technology have improved image quality dramatically giving the user the ability to identify even subtle abnormalities. New(ish) modalities such as tissue doppler imaging, speckle tracking and 3D reconstruction have allowed comprehensive assessment as well as opening new possibilities for interventional procedures such as Transcatheter edge-to-edge repair of the mitral valve.

Piezoelectric Effect

Ultrasonic sound waves are generated by the probe via the piezoelectric properties of certain ceramic materials located within the probe itself. This effect was originally described in 1880 when quartz was used. Piezo-electric materials deform mechanically when an electrical current is applied to them. This deformation can be controlled via an applied electrical current to oscillate at a resonant frequency generating a sound wave.

High frequency sound waves exist in the “ultrasonic” range and for most probes this is in the 5MHz range. The machine can alter periods of compression and rarefaction within the ultrasound wave. The overall pattern of oscillation is in parallel to the direction of the beam (unlike waves in the sea which are transverse to the direction of travel) This concept can be confusing as many diagrams (including the one below) show sound waves represented in a transverse fashion which is technically incorrect.

The Ultrasound Wave

Sound waves are emitted in a longitudinal manner where oscillations of compression and rarefaction progress in parallel to the wave direction  (unlike ripples on a pond which oscillate at right angles).

Once emitted from the transducer the ultrasound wave embarks on a journey first through the “matching layer” on the ultrasound transducer, next through the ultrasonic gel medium and next through any tissues, bony structures etc... that it may encounter within the patient. On it’s journey the ultrasound wave can undergo one of the following:

1. Reflection
2. Refraction
3. Scattering - this determines the “texture” of a given structure in the ultrasound image.
4. Absorption - energy lost as heat

Reflection

Reflected sound waves are interpreted and processed by the ultrasound probe to create the ultrasound image. The probe is able to detect how long the wave has taken to “return to base” and cause a further mechanical deflection of the piezoelectric crystal resulting in a voltage. Voltage signals are then processed by the echo machine to determine the depth at which the reflection occurred (derived from its time taken to return) as well as the amplitude of the reflection (how “strong” the reflection is).

Highly reflective structures tend to be those containing calcium such as bone or calcified vessels; these structures appear a bright white on the ultrasound image. Tissue reflects some of the ultrasound wave energy but also absorbs and scatters some resulting in a grey (of varying shade) Air-filled structures do not reflect any ultrasound waves and appear black.

The average propagation speed of ultrasound waves in soft tissue is 1540m/sec. This has practical relevance as increasing the frequency of the probe, say to 10MHz improves axial resolution by halving the wavelength of the beam, this is at a cost of decreased penetration. This explains why intracoronary ultrasound used in “IVUS” (which only has to travel a few millimetres from the source)  operates at 20Mhz or higher to improve axial resolution. By the same token “low frequency” ultrasound probes are used to image deeper abdominal structures.

Reflection of ultrasound signal occurs best when structures of interest are “in-line” with the ultrasound beam and reflection occurs back towards the probe. Structures at the peripheries of the echo sector scan window are often poorly visualised, the sector scan should therefore be narrowed as far as possible (within reason so that the echo operator can maintain orientation) whilst trying to centralise the object of interest.

Refraction

As the ultrasound beam moves between tissue planes of differing densities the incident angle of the beam will also change, this is known as “Snell’s Law”. The two major factors affecting this change in angle are the incident angle (angle upon entering the tissue plane interface) and the change in velocity of the sound wave between the two tissue mediums which in turn relates to differences in density between the two tissues.

Frequency

Frequency is what determines the “pitch” of the sound i.e. high frequency sounds tend to a have a high pitch. Ultrasonic waves are inaudible to human ears but the same principle applies. Frequency is measured in “cycles per second” which are known as Hertz (Hz) with 1Hz being 1 cycle per second. It is worth noting that as the frequency increases (i.e. number of peaks per unit time increases) the time period decreases.

Using a surfing analogy the frequency will be the number of waves that pass underneath you on your surfboard in a given period of time. Most transthoraic ultrasound probes operate in the 2.5Mhz range with intravascular ultrasound (IVUS) used by cardiology in the cardiac catheterisation laboratory in the 20MHz range as the depth of penetration required is so low; only the thickness of the arterial wall.

Time period is the time taken for two successive reference points on the wave to pass an arbitrary but fixed point in space. Frequency is therefore the inverse of the time period of a wave.

Time Period

This is the time taken for two identical locations on a given wave to pass an arbitrary position in space. It is the reciprocal of frequency and is described in units of time (e.g. milliseconds) Using a surfing analogy it can be thought of as the time you have to wait on your surfboard until the next wave comes along (to pummel you onto the seabed!)

Bandwidth

Bandwidth refers to the range of frequencies that the probe is able to transmit and interpret. Various filters can be placed on bandwidths to be analysed, for instance focussing on only “high velocities” when examining turbulent blood flow or “low velocities” when using tissue doppler analysis and filtering out frequencies outside of this range.

Wavelength

At any fixed time the wavelength is the distance between two identical corresponding points on successive waves i.e. the distance between two peaks. This is measured in units of distance e.g. micrometers. Frequency and wavelength are related by the speed of sound within the medium. In tissue of “average” density this is 1540m/sec. Using a final surfing analogy this is the physical distance from the wave you are currently duck diving underneath and the next one in the set appearing behind it about to hit you on the head. Waves with a longer time period will therefore have a longer wavelength.

Propogation speed through biological tissues is quite uniform only varying slightly with different tissue types. The main  factor determining acoustic impedance to the ultrasound wave becomes the density of tissue that it is travelling through. The wavelength for a standard 5Mhz ultrasound probe is approximately 0.3mm

Anatomy of the Ultrasound Probe

The echo probe itself contains four main components which are required to generate an image for interpretation. These components are all fixed at the point of manufacture but remain a popular topic for exam questions to show candidates understand the principles underlying image acquisition.

1. Backing material

Provides a damping effect to oscillations created by the piezoelectric crystals. This prevents the crystals from “ringing” like a fingertip rubbing over a wine glass. This effect serves to increase the bandwidth of frequencies that the probe can emit which improves axial resolution for shallower structures of interest. The enhanced damping provided by the backing material serves to decrease the “Q-Factor” (see below)

2. Piezoelectric crystal array

These are ceramic crystals which exert unique electromechanical  properties conferring them the ability to transduce electrical into mechanical energy. The piezoelectric effect is derived from the Greek “Piezen” meaning “to squeeze” referring to the mechanical contraction and expansion of the crystal that occurs when an electrical current is applied.

This unique property results in oscillation at a certain (ultrasonic) resonant frequency when a given electrical current is applied to the crystal (and vice versa). This resonant frequency is determined by the crystal material as well as the thickness of the piezoelectric crystal layer. The crystals can be “stimulated” with an electrical current resulting in the generation of an ultrasonic wave. They can also “detect” ultrasonic waves reflected from the tissues creating mechanical deformation resulting in a small electrical current which can be detected by the echo machine.

Quartz (silicon dioxide) is the crystal classically described as having piezoelectric properties and zinc oxide was used successfully in early “crystal radios” which work by harnessing the piezoelectric effect. Most ultrasound probes use lead zirconate for the piezoelectric crystal layer.

3. Matching Layer

This layer serves to optimise the transmission efficiency of ultrasound waves into the tissues as well as facilitating their optimal transfer back into the piezoelectric detector crystals. There is a large difference in impedance from the ultrasound source (33MRayls) to the tissues (1-3MRayls).  The overall purpose of the matching layer is to optimise energy transfer into the tissues by gradually decreasing impedance in a stepwise and controlled manner.

This layer results in an overall energy transfer of around 90%, much of which would otherwise have been lost in a large step down in impedance from the piezoelectric crystals to the tissues without it. Optimal thickness of the matching layer is 1/4 of the operating frequency of the transducer elements, this is extremely thin and is one of the more challenging parts of the probe manufacturing process. Multiple matching layers may be used in some transducers. Thicker matching layers are cheaper but increase attenuation and therefore result in poor signal quality.

4. Acoustic Lens

This serves to attenuate beam width as well as focussing depth. The lens is made from a specialised rubber which has low impedance and matches well with body tissue velocities (around 1000m/sec) It serves to focus the ultrasound beam and improve image quality by reducing scatter.

The above information outlines the components of the simplest possible form of ultrasound “probe” which only allows imaging along one individual scan line. Modern probes are organised into an “array” of piezoelectric crystals which work synchronously to create a real-time image across a “sector” of a given width. TOE machines use a “phased array” which refers to the way in which adjacent groups of crystals are “activated” after very short time delays. Time delays allow for angulation of the beam away from the perpendicular face of the probe improving the width which can be scanned.

3D-capable TOE probes utilise a “matrix array” of crystal units within the probe tip allowing for a pyramidal sector with depth, width and also elevational width characteristics, all of which can be adjusted and optimised.

Ultrasound Beam “Physiology” and Characteristics

Constructing an image for interpretation requires a pulse of ultrasound to make a “round trip” from the transducer, to reflect off of the object of interest and then return back to the transducer for analysis. In order to avoid range ambiguity (not knowing how far away a given  signal is) this whole round trip process must be completed before the next cycle can start. This round trip takes longer if objects of interest are further away from the probe and this is one reason why deeper structures are often poorly visualised using ultrasound.

The above”round trip” describes the journey of one pulse on a single scan-line of the eventual 2D sector image. Multiple scan lines at different angles are used to create a 2D sector image to a depth and width specified by the user. The “standard” width sector will use 128 scan lines. Overall the depth and width should be the minimum required to examine the object of interest whilst maintaining orientation with adjacent structures in order to optimise the image resolution. The different types of resolution and how to optimise them are described in more detail below.

Pulse Duration

This describes how long the individual pulse will last for (in microseconds) and is typically only for a few cycles. The waveform within an individual pulse can be emitted at different frequencies and amplitudes. The pulse duration (in microseconds) corresponds to a pulse length (PL)

Pulse Repetition Frequency (PRF)

This is the number of “pulses” emitted by the probe in a given period of time (one second) and is expressed in “cycles per second” or Hertz (Hz). Pulse repetition frequency is determined by the speed of sound (1540m/sec in tissue) and the distance that the sound wave needs to travel. The speed is fixed with the body and so PRF varies in relation to depth with lower frequencies being required at greater depths as more time is required for the sound wave to cover the greater distance. Each pulse generated and detected gives a “snapshot” of what is happening at a given point in space. It therefore follows that the more “snapshots” that the machine can record, the more accurate the image, shallow structures can be assessed with more pulses and therefore be viewed with better resolution. This applies specifically when considering myocardial velocities (tissue doppler) and blood velocities at a specific point (pulse wave doppler - see below)

In practical terms, adjusting the pulse repetition frequency (PRF) results in the machine adjusting it’s in-built “wall filters” which analyse returning signals of a certain frequency.

Duty Cycle

This refers to the ratio of “on time” i.e. the length in time of the pulse in comparison to the “off” or listening time. Short duty cycles mean that the ratio is small and the PRF is likely to be high. This results in more energy being transferred to the tissues per unit time, thereby increasing power.

Amplitude

This describes the wave “heights” when compared to the average across all waves. High amplitude therefore suggests a large deviation away from the mean value. Using different amplitudes results in different “shades of grey” on the resultant image on the ultrasound machine.

Quality “Q” Factor

This is a term used in ultrasonic physics referring to how much damping an emitted signal from the piezoelectric crystal undergoes. There is an exponential decrease in signal oscillation as time passes, the rate of this decrease is referred to as the “Q-Factor”. The backing layer of the ultrasound probe serves to promote damping thereby reducing the Q-Factor i.e. improving damping.

Ultrasound Adjustment and “Knobology”

This section describes the different ways that the sonographer can adjust parameters on the ultrasound machine in order to focus on different areas, improve image quality or measure certain characteristics. These take practice to implement but are worth persisting with as the final images will be easier to draw clinical conclusions from.

Power

This is the amount of energy transferred per unit time and is given the SI unit of watts. The “duty cycle” and the amplitude of the emitted waveform determine the power. Power can also be measured with reference to an arbitrary “reference power” and helps with the large differences in orders of magnitude that are often required when using ultrasound. This reference power is compared to the relative power being emitted by the probe and this ratio is then expressed on a logarithmic scale in decibels to make comparison between different power outputs more straightforward.

In most machines the “transmit power” can only be turned down from a safe maximum which is predetermined by the manufacturer. This is normally only of relevance if the TOE probe becomes hot and you still wish to image.

Intensity

This is power divided by a unit area (the surface area of the ultrasound probe) and plays the most important role in biological damage caused by the ultrasound probe.

Gain

Gain is often confused with power as increasing both can make the image look “brighter”. Gain is a post-processing feature which is applied after echoes have returned to the transducer (whereas power concerns the initial energy required to generate the sound waves in the first instance). All returning amplitudes are increased by a certain percentage (according to how much gain you set) resulting in an overall “brighter” image. Note that areas of the image which were previously dark also brighten so delineating differences between structures is not improved.

Time-Gain Compensation (TGC)

This concept is sometimes referred to as “depth-gain compensation” and refers to progressively amplifying returning signals from deeper structures in order to maintain a uniform image that doesn’t “fade” as you look deeper. Acoustic attenuation occurs as ultrasound waves travel further into tissues due to internal friction, scattering within non-homogenous tissues and reflection. This attenuation effect is exponential with respect to depth and also depends on the tissue-type being imaged; blood and fluid-filled cavities having lower attenuation coefficients compared to bone and cartilage which have high attenuation coefficients.

In practical terms this automatic compensation can be finessed further by adjusting individual “TGC-slider dials” on the echo console (or an electronic version at the side of the screen on modern TOE machines). This should be focussed over the area of interest to optimise image quality i.e. don’t slide the bar down to a deep structure when examining a superficial one.

Lateral Gain Compensation (LGC)

This is a similar concept to TGC where structures at the periphery of the sector scan can have their gain either increased or decreased as required. The typical example for where this may be useful is an image containing a blood-filled ventricle in the centre with bright echoes of adjacent tissues being seen laterally. The gain for these peripheral structures can be reduced (or increased) as required by the user. In reality this is a little-used feature and rotation of the probe to the area of interest or narrowing the sector width often improves image quality

Dynamic Range

This refers to the range in amplitudes between the highest (or brightest) signal and the weakest (or dullest) signal and can be considered as the “Greyscale” available for assigning to pixels on the final image. The dynamic range of most modern ultrasound transducers is in the region of 60dB and allows for identifying differences in tissue-types and subsequent characterisation of these structures on the final image. A wide dynamic range allows for discrimination of subtly different structures within a homogenous tissue-type. One possible application of this in echocardiography would be examining the myocardium of patients with infiltrative diseases such as amyloidosis and sonographers typically may increase dynamic range when imaging the liver looking for subtle abnormalities.

In most cases the “full” dynamic range of 60dB is compressed down to 20dB or so which loses some of the subtle differentiation between tissue structures. Reducing dynamic range results in a less detailed image but with higher contrast.

Compression

This is another post-processing feature where the dynamic range of returned signals are compressed together. Having a smaller dynamic range means that the largest returning signal remains the largest and the smallest remains the smallest but the overall range displayed is reduced. Compression is the inverse of dynamic range and the usual compression in echo is down to 20dB (from a full dynamic range of 60dB). Compression loses some of the ultrasound’s ability to differentiate subtle differences in tissue types resulting in fewer “shades of grey” but does exaggerate differences between obviously different structures (similar to improving contrast on a photo). Compression can be adjusted manually by the user but is frequently set to a machine-generated “optimised” setting which balances tissue differentiation with improved contrast.

Resolution

Resolution is the ability of the echo probe to distinguish nearby objects in space, returned signal strength or time as separate from one another. It is considered in two main domains: spatial and temporal. Spatial resolution in turn is divided into axial and lateral resolution and is largely determined by scan line density whilst temporal resolution is determined by the frame rate of the recorded ultrasound loop.

Axial resolution

This describes the ability of the ultrasound probe to differentiate two objects of interest which lie superficial and deep to one another as distinct entities. It is largely determined but the spatial pulse length which in turn relates closely to the depth of the structure being imaged. Overall axial resolution is improved when examining shallower structures and depth should always be kept at the minimum required ensuring all objects of interest are kept in view.

Lateral resolution

This describes the ability of the ultrasound probe to differentiate between two objects of interest which are side-by side one another on the image. It also explains the loss of resolution at the edges of the TOE sector scan which are mostly due to reflection of echoes away from the transducer which increase as you move laterally. This effect can be minimised by reducing the sector width to only focus on the structure in question and centralising the object of interest. Lateral resolution is also known as azimuthal resolution.

Temporal Resolution

This refers to the frame rate of the real-time image being displayed on the ultrasound screen. Frame rate is determined by sector width, sector depth and also elevational height when using 3D imaging. Temporal resolution falls away dramatically with 3D imaging due to the physical limitations of having to wait to “listen” for reflected echoes across multiple scan lines across multiple thicknesses. This temporal resolution problem can be overcome to a point by using “multi-beat” acquisition which uses computer software to artificially “stitch” together 2-4 individual narrow sector volumes together to acquire an overall image. The quality of this is degraded by movement artefact and so asking the patient to hold their breath or doing a breath-hold on the ventilator is one way to minimise this effect.

20 frames per second (FPS) is the lower limit at which the human eye can detect “stuttering” of the image but one should always aim to try and optimise the frame rate as high as possible when scanning. A high FPS means when recorded loops are slowed down for detailed post-processing and analysis temporal detail can be maintained.