Ultrasound imaging had its genesis after the end of the second world war, when a few bright sparks with nothing better to do realised that if you took a surplus radar display, coupled it to a high frequency ultrasound transducer rather than a radio antenna and then put the transducer in contact with the body, the radar display was pretty well calibrated to register echoes of the ultrasound reflecting off structures inside the body. And on such experiments, a new branch of medical imaging came to be built.

The key elements?  A surplus of rather expensive (and big) hardware going cheap; compact transducers suitable for contact with the body surface; some wiring to connect them.  Lots of twiddly knobs on the hardware were essential because the operator had to be able to tune the circuits in order to get the oscilloscope to show those echoes.

It didn’t take long for the radar display to be ditched for something purpose designed, but that box was still pretty large.  So it was easier to bring the patient to the box, and have a number of different probe designs for looking at different parts of the body.  Bear in mind that ‘real time’ ultrasound was only being used for things inside the body that tended to move around – babies, heart valves and the like – which would be tracked using M-Mode (movement mode) while structures that didn’t move much were displayed by moving a gantry-mounted probe over the body surface, building up a picture of what lay below.  Patients were covered in a slippery coupling medium (olive oil was good), and asked to hold their breath and stay absolutely still while the operator built up the image. Lots of tuning was still required to get that oscillograph to display the echoes right.  Also because a beam of ultrasound has a tendency to spread out, and high frequency gives better definition but less penetration, there were a large selection of probe heads to choose from, each having a different frequency and focus.  A head with the right combination of focus and frequency would be screwed into the gantry housing to display optimally the structures of interest.  The process of scanning was quite a skill with many choices governing the optimisation of the equipment.   And so developed the role of the sonographer.

Then several people had bright ideas:

• Can we put the image onto a TV screen? Yes we can, using a nifty new piece of electronics called a Digital Scan Converter which maps the ultrasound lines onto TV scan lines.  Bin the oscilloscope.
• Can we put wheels on the scanner to make it mobile? Yes we can, and you can even put the TV display on a nice arm so that it can be in a more convenient position.  But don’t make the TV too big because it’s still quite heavy!
• Do we really need this gantry thing? No we don’t.  If we use a real time probe, we can use B-mode (brightness mode) to show us a 2D picture, and if we move the probe around on the body surface our brain can learn to put together the real time picture into a meaningful map, and then we can zone in to scan the critical structures in detail. More skill for the sonographer to learn.
• Can we get rid of the need for so many different fixed focus and fixed frequency probes? Yes we can.  If we borrow the concept of a new type of scanning from our radar friends, we can make a transducer array. This can give us electronic control over focus and even frequency, thus making one design of probe more capable of scanning different sizes of people and different structures within them.  Of course that’s as long as the operator sets the controls right!

By the mid 1980s things were getting pretty sophisticated, and a cardiac surgeon in Japan asked some engineer friends of his if it was possible to make a real time picture of blood flowing inside the heart and great vessels.  He wanted to see the jets of high velocity created when blood was forced back through an imperfectly closed heart valve as these were quite difficult to spot, and also see when blood was being accelerated through a narrowing in a vessel.  The engineers came up with a display format where blood flowing towards the probe was coloured red and blood flowing away from the probe was coloured blue.  They injected green, to make white pixels if there was very high velocity turbulent flow such as you’d be likely to get in a regurgitant jet.  Presto! Colour Flow Doppler is born.

During the 1990’s and into the 21^st century faster more powerful computing allowed designers to start improving the image by using and reusing information.  This was initially important in the development of Colour Doppler as a usable tool but soon the same techniques were being adopted to create better B-mode images.  It accelerated still further when good old Moore’s Law had progressed enough for manufacturers to ditch the beamformer (a rather elegant analogue technique for assembling echo data) and replace it with the digital version which allowed much more flexible manipulation of echo data using lots of memory and fast, powerful microprocessors.  This gave rise to a whole raft of new controls and technologies with complicated and whizzy-sounding names and acronyms – harmonics, speckle reduction, persistence, compounding etc etc – which could be cooked together to make better quality images

With such complication, the task of the operator, like a chef assembling all the ingredients in the correct proportions for a harmonious end product, has become ever more difficult.  Manufacturers have enhanced the opportunities for creation of purpose specific presets and employ an expanding staff of expert applications staff who work to optimise scanners for a particular task.   Their essential support role for the user is covered within the prices charged for the goods.  Maintaining the complexity of ultrasound scanners has helped to support prices in a self-sustaining cycle.

The loser is the practitioner who wishes to use an ultrasound scanner for one particular purpose. Ultrasound in a community/out of hospital setting for examination of musculoskeletal structures is one such example.  High performing systems for MSK scanning are large, expensive and multitalented, so likely to be found in the main imaging department of a hospital, being used by sonographers and radiologists for a wide range of tasks.  Access for the MSK specialist may not be easy.  These are also expensive tools for the independent practitioner specialising in MSK scanning with capabilities extending beyond what is needed.  It’s true to say that the vast majority of requirement can be met with less sophisticated equipment, and since these systems are often compact, they also offer mobility which larger systems lack and finally they offer better value for money.

There remains a challenge.  These systems are designed by the same people who put together the large expensive machines, so invariably they follow similar design themes; complexity among them.  The independent practitioner still needs to know what controls affect image quality and which ones should be most to the fore, so this is a short reminder:

Basic Controls:

Image brightness

B-gain.  Usually the largest knob on the control surface and not without reason.  It’s analogous to the volume control on a radio and makes the overall image darker or brighter.  Just like turning up the volume, weaker echoes will begin to be displayed, but the strongest echoes may start to saturate if the gain is set too high.  Always be aware that our eyes reset themselves for the overall brightness of the surroundings so if you do not compensate by increasing the gain in a bright room, you will miss subtle textural detail in the image.

Dynamic Range.  This control is normally in secondary controls, but works in tandem with the B-gain.  The higher the dynamic range number, the greater the spread of echo strengths between black and white.  High dynamic range flattens the image, making it less contrasty.  Each person will have their own preferences, but it is important when scanning in a light room to ensure the dynamic range is turned down to give higher contrast as we lose the ability to resolve subtle greyscale differences in bright light.

Output Power.  This is equivalent to how loudly the probe shouts when sending out its pulse.  A stronger, louder pulse will generate stronger echoes and that is generally a good thing, except the louder sound bounces around a bit more inside the body and that can create more artifacts, reverberations and distortions.  If you’re scanning something superficial, clearer images can be obtained by turning the output power down and the gain up.  Normally output power control is a tertiary control, but it is worth knowing where to find it and if you routinely scan superficial structures, it may be worth setting a preset with lower output power, particularly if you are using a lower frequency probe.

Focus, Frequency and Harmonics

Focus.  Setting the focus is important.  Normally there is a small arrow to the side of the image which shows you where the focus is.  Make sure it’s at the same depth or slightly below the structure you are most interested in because that ensures it is near the point of highest lateral resolution.  Above and below this area, there is a progressive fall away of resolution and the beam is more spread out.  A longer (deeper) focus requires a larger aperture; the penalty of large aperture is loss of definition in near field.  Shortening the focus makes a small aperture, which is fine when the structure you’re interested in is superficial and close to the probe, but the beam then disperses quickly and you will find deeper structures are poorly visualised.  Rule of thumb – moving the focus deeper in the image increases penetration and allows you to see deeper structures.  If a scanner does not show the focus with a marker, then by convention the focus is automatically around the middle of the image and the operator must change the depth to ensure the structures of interest are in the middle section of the screen. If focusing is not done automatically then ensure the focus control is conveniently placed because you should be using it.

Frequency.  Setting the frequency is important.  All systems these days actually use broad bandwidth, which means that they get the probe to emit a pulse containing a range of frequencies and the probe then listens to a preprogrammed pattern of echo frequencies that starts high close to the probe and finishes low at the bottom of the image.  This optimises the balance between resolution and penetration and helps make probes more versatile.  Altering the frequency changes two things; firstly it shifts the balance of frequencies in the outgoing pulse a bit and secondly the preprogrammed pattern of echo frequencies being listened to is tweaked (eg listening to higher frequency at each point in the image, or lower frequency at each point in the image).  To use a musical analogy, this is like getting a different note out of an instrument according to the way it’s played.  The basic design of the instrument sets the range of notes it can play (Piccolo vs flute vs bass flute) the variety of different notes played in that range is down to the design and natural resonances, then the player plays the tune!

Harmonics.  The use of harmonics is a step further.  Ultrasound does not simply pass through the body, it interacts with the body tissues and is changed by them.  Building an image is largely done by predicting what the ‘standard body’ does with ultrasound and choosing to listen at the right frequency with the right amount of amplification.  However, people are not all the same, and some body types seem to interact unpredictably with ultrasound which results in a degraded image.  Scientists investigating this discovered that these difficult to scan body types were transferring energy from one part of the frequency range to another part of the frequency range, typically a harmonic of the original frequency (eg signal going into the body at 4MHz would generate echoes at 8MHz even though there was no outgoing signal at that frequency.  The echo signal at 4HMz was as a result weaker than expected).  This is a bit like trying to create low notes on a clarinet – if you blow too hard, you get a squeak because the energy flips into a higher harmonic.  By programming to listen to the harmonic, you could actually improve image quality and, as manufacturers have learned to utilise this phenomenon, harmonic imaging has become more and more popular.  Be aware however, that harmonic imaging and ‘fundamental’ imaging (without harmonics) do need to be used according to the body type of the person being scanned and how the machine is playing its tune, so this control is one you have to be prepared to use actively.