Self Assurance to Quality Assurance: Knowing the Correct Things to Do
In this chapter, we will address those issues that are very important to professional sonographers, namely, the ability to be sure that you are doing a good, precise, and reliable job. The ability to know that the quality of work you produce is of the highest nature not only requires training, but also an understanding of the technology and instrumentation as well as issues that pertain to overall quality assurance topics. We have taught physics to all grade levels and will use some of the simple techniques to explain the phenomenon of ultrasound. We will then expand upon this information to paint a picture that will allow the concerned ultrasonographer to be able, on their own, to assess the quality and reliability of their work.
Fun with Physics
In order to get a simple mental picture of what is happening with your ultrasound instrumentation, we would like to describe a situation that we all have experienced. Envision in your mind a large field. Directly across from you on the other side of this field is a very large cliff. Not only is this cliff tall, but it is very sheer; the surface of the cliff is quite smooth up and down but not left to right. Now, if you were to let out a yell (Hey!!!) and wait a short period of time, you would undoubtedly hear the echo of your voice come back. If you did this over and over again, you would quickly realize that the time it took for your voice to get back to you after you yelled was the same in each instance. Clearly, the sound is leaving your mouth, traveling across the field, hitting the cliff and coming back again. Let’s suppose that we get a little more quantitative and use a watch and discover that, each time we yell, it takes a full two seconds for the sound to come back. Obviously, it took one second for the sound to get to the cliff and then an additional second for the sound to come back. For discussion purposes, assume that the speed of sound in this air is 1000 feet per second. (Although we know that the speed of sound varies with humidity and atmospheric pressure, this is a pretty good number to work with.) Since it took two seconds for sound traveling 1000 feet a second to get back to us, the sound must have gone 2000 feet round trip. We can then conclude that the cliff is 1000 feet from where we are standing. This is exactly the same way that ultrasound can be used to determine distances to an object. In this particular case we are using our voice. With ultrasound probes we use much higher frequencies which are better at getting through tissue and liquid.
We can develop this scenario further and assume that we have one hundred of our fiends standing shoulder to shoulder facing this cliff. We give each of them a watch and have each of them sequentially yelling out “Hey!!! ” at the cliff and measure how long it takes for the ehco to come back. When we look at the timing information from each of these hundred people, we note that the number of seconds that it took for the sound to come back to each person was slightly different. This would give us an indication that each person was slightly farther away from, or closer to, the cliff. By looking at the time that it took at each position, we could get an understanding of the roughness of the cliff. This is identical to using an ultrasound array transducer.
Take things one step further. Imagine that somehow or other this cliff is moving toward us as we stand in this field. In this instance, instead of yelling out “Hey!!!,” we take a musical instrument like a flute and play a pure note, say a middle C. We play for approximately one second and then wait and listen for the echo to come back. Lo and behold, when the echo comes back it’s a different tone; instead of a C, it’s a D, a higher frequency than we sent out. We know that sound bouncing off of a moving object will have its frequency changed. A higher frequency would mean the object is moving towards us, and a lower frequency would mean the object is moving away. Observing a different note coming back to us prompts us to use equations developed by physicists to determine how fast that cliff is moving toward us. The larger change in the note, the faster the object is moving. This change in the note is known as the Doppler effect.
What we have described is what happens in the ultrasound clinic. Measuring the time for the echoes to come back tells us how far different objects are from us. Measuring the new note coming back allows us to calculate how fast that object is coming toward us or moving away, depending on whether the new note is of a higher or lower frequency. The fundamental theory of clinical ultrasound measurements is no more complicated than this. It is simply a clever utilization of different frequencies to deal with the fact that we are not just yelling through air, but we’re yelling through blood or tissue or even bone.
One other real life example, which is found in almost every Doppler ultrasound book, is the analogy of standing by a railroad train track as the train is coming toward you blowing its whistle. You will hear a certain note coming from the train as it heads face on toward you (making sure you are standing at the side of the tracks, safely protected). At the moment that the train passes where you are standing there is a sudden lowering in the note or the frequency of the sound. This is that transition point where, instead of the sound being brought to you on a carrier, after it passes the point where you are standing it shifts to a lower frequency because the carrier (train) is moving away.
We now have the ability to tell distance from us, referred to as the B mode or imaging mode of ultrasound. We also have the ability to determine if something is coming toward us or going away from us and how fast it is moving by using the Doppler measurement. Exactly how the moving object changes the frequency is not really relevant for this discussion It is just important that we understand, right now, that an object moving towards us change the note that comes back.
What about ultrasound probes? Are these more complicated than just simple sound generators? The question is a good one, and the answer is no. Ultrasound crystals are simply tiny speakers. They emit sound as fluctuating voltages are fed into them. Yes, of course, they are quite elaborate and made of wonderful materials, but so are regular stereo speakers. The fact is, ultrasound crystals can be viewed as simple speakers in a stereo giving off sound. The unique thing about these ultrasound crystals is that you can also use them as a microphone. In other words, at one point, you are using them to give out the sound like a speaker. Then you wait for the sound to come back, at which point they function as microphones. Although we will be dealing with frequencies that are very different than we normally hear in our stereo or our radios, it is the exact same phenomenon and there is no reason to be confused at this physical phenomenon since we all have experience with stereos, radios, speakers, and with our voice echoing off of cliffs.
So let’s quickly review the simplicity of an ultrasound exam. The ultrasound probe is like a little speaker that’s giving off a sound wave which goes into the body, bounces off certain objects, and comes back out of the body. The amount of time it takes to come back is a function of how far away the objects are and the amount that the note is changed is a function of the velocity of the object that the sound is bouncing off.
There are also issues related to the amount of sound that’s absorbed by the different tissues, which gives us some imaging discrimination, but in general, there is no fundamental difference in the process of sending out the sound and having it bounce back to you.
Now we seem to be experts in the physics of ultrasound and we’ve perhaps attended a sonographer class. But when we sit down in front of these instruments with the multiple computer processing capabilities and the numerous controls (which we refer to as knobology), things seem to be a bit more daunting to us. They need not be. The trick for any technologist using an instrument is to be able to train on that instrument. In the long run, one has to train on a human patient, but until the technologist is totally comfortable with many aspects of tuning the instrument, it would be very beneficial to be able to train on an artificial system.
There are two types of artificial systems available that we will discuss them very briefly here. The first is an imaging phantom. Imaging phantoms are basically little boxes that contain gel with ultrasonic properties similar to human tissue. Imbedded in this gel, at different depths, are different plastic or metal objects. You can sit beside one of these boxes and become familiar with the adjustments necessary on the instrument to be able to find these hidden objects. Imaging phantoms are reasonably priced, fairly easy to build, and easy for the ultrasonographer to use. When given an opportunity to sit in front of a complex ultrasound instrument for extended periods of time, it gives the ultrasonographer the freedom to practice and carry out exercises that will improve the ability to capture reliable and reproducible results. It will also give you, the sonographer, the confidence necessary to go ahead and start testing on human subjects.
There are several companies that sell these phantoms (1-3). Although there are claims that some of them last longer than others, they all seem to do a pretty good job for the sonographer who is in training.
The other aspect of ultrasound testing that a sonographer needs to be able to practice on is the use of the Doppler system. Clearly, the utilization of Doppler is the highest form of technology that the ultrasound instruments can deliver. It is true that imaging is very neat to look at. It is very important to find objects and check dimensions, but the utilization of Doppler to noninvasivly evaluate fluid flow in the body is a remarkable accomplishment.
So what we need to do is develop a phantom system that allows the technicians to practice finding and accurately determining velocity profiles. Here is where some of our previous discussions come into play. Doppler is the measurement of velocity. This means that Doppler measures how quickly something is coming towards the probe. If something is coming at a slight angle to the probe, then there are software steps in the machine that account for this. The important issue here is that any device that we want to use to train with must have a known and controlled velocity within it. Beware of flow phantoms that simply deliver a certain amount of volume per time. A device that can, for example, deliver one liter per minute without telling you specifically the velocities that are built into it is of little use in determining whether or not you, as a practicing ultrasonographer, are able to find accurate velocity numbers. It is true that most ultrasound machines have the ability to print out flow results, but this is a mathematical calculation that takes the velocity information from the probe and processor along with the dimensions of the vessel under investigation and calculates out a flow rate. Doppler is not a flow measurement, it is a velocity measurement and if there is nothing else that you take from this chapter, that’s the most important issue. Once you realize this, then you can more properly evaluate flow phantoms or velocity phantoms to be used for training. Purchasing an inexpensive flow phantom that does not give you velocity information is senseless.
The complexities of finding flow within a phantom that has defined velocity has certain training implications that cannot be addressed by simply sending an electronic signal into the probe. For this reason, most quality assurance individuals involved with ultrasound physics do not recommend the use of electronic devices for training or calibration assistance.
This leads to a basic question: Which Doppler phantoms can be used for sonographer training? One approach is the use of the so-called string phantom (4). In this device, a string with a target on it is immersed in a tank, similar to a fish tank. The string is moved between two hubs and the probe focuses on the little target in order to get velocity measurement from the target as it comes toward or away from the probe. String phantoms can be useful in that they are one of the few commercially available objects that tell you what the velocity of the target is. By knowing this, you can compare the numbers to the velocities that your instrument is showing. It also allows for the measurement through various liquids so as to more appropriately replicate the acoustic properties of the human body.
On the downside, the string phantoms represent a somewhat messy situation in that this fish tank has to be full of liquid. There have been reports of string phantoms whose indicated or theoretical velocities were measured by monitoring the speed of the hub upon which the string rides. In those instances in which the string slips on the hub, the displayed velocities are simply wrong. The problem with this set-up is that a user cannot tell by looking at the instrument that it is displaying an incorrect velocity. In addition, there are complaints about the material of this moving target and the intensities and reflections that come back from it being so different from anything that one will investigate in the body that it is rendered much less useful. David Evans in his textbook on Doppler ultrasound (5) states that a proper Doppler velocity phantom should be pulsatile, contain small reflectors similar to blood cells, and pump liquid with acoustic properties similar to that of blood.
This brings us into the realm of the actual flowing velocity phantom. A properly designed velocity phantom should cover all the velocity ranges that can be acquired in the vascular system under most circumstances. A government panel recently put together a “wish list” in which the velocities of interest ranged from 35 to 500cm per second. We know of applications in which even slower measurements would be of interest, but these are more in the realm of research. With a properly designed flowing velocity phantom, an ultrasonographer can sit beside this signal for several hours at a time, practicing on a brand new ultrasound instrument, utilizing all the different control systems to adjust intensity, contrast, etc., and do this without the problems of constantly varying signals that come from patients. I have been told repeatedly by ultrasonographer friends and colleagues that it is very difficult to train on human patients because subtle changes in the setting of the knobs are overridden by the very large changes associated with patients that are moving, breathing, or sweating.
As an ultrasonographer, you should have the opportunity to privately and independently practice on any instruments or any probes on which you feel you need additional practice. And you should not be compelled to do this practice on patients. There are commercially available devices (6) and they should be made available to all professionals so that you yourself can be assured that you know what you are doing with these instruments.
Over the past couple of years, we have been involved in a number of studies of ultrasound processors, probes and technicians. In every case, the technicians have expressed a desire to utilize the instrumentation that we had available to practice with because of their concern for a lack of training time that was unencumbered by the need to practice on humans.
In the previous section, we reviewed the need to have mechanical testing devices or training devices that allow ultrasonographers to practice until they are comfortable with an instrument that they are required to generate medically important data on. There is also a need on the part of hospital administrators, laboratory managers, professional societies and other training institutions to test an individual sonographer for the ability to accurately measure dimensions and velocities from an experimental system. Again, appropriate imaging and flowing velocity phantoms can be utilized in this case. It is clear that if an individual cannot go to the velocity phantom and measure the correct velocity that it is set to, then it is not time for them to go and test on a human. This gives training and testing professionals the ability to quantitate an individual’s certifiability. As well, if these individuals in training are allowed to practice for extensive periods of time then they can confidently pursue their certification testing as well as begin work on humans.
These discussions obviously lead to the issue of quality assurance. There are four major components to a successful ultrasound laboratory. First is the actual ultrasound processor itself. This is the big box with a screen on it and all the knobs and buttons. Next is the probe. These ultrasonic probes are fantastically complicated and are very susceptible to damage by dropping and aging. A perfectly functioning processor will not produce valid information if the probe is damaged. Third is the recording system. Most sonograms are read by physicians after the fact. If the recording system used to store that information for further review is in error or of poor quality, then the physician will receive poor data. And, last but not least, is the sonographer, the fourth component of a successful laboratory. A professional technologist, in most instances, must be willing to spend as much time training as necessary to do a good job. Unfortunately, many sonographers make mistakes because they are working with systems that they cannot check. In other words, mistakes are being made by technologists who think they are doing a good job. Because they have no way of testing their instruments, probes, or themselves, they are, in some cases, producing erroneous results.
We published the results of an experiment that we did (7) and I have reproduced some of the data. Figure 1 is an extensive compilation of the data from fifteen hospitals testing over fifty ultrasound instruments. As you can see, some of the errors are shockingly high. I must stress that these were ultrasound instruments and technologists that were functioning and operating in licensed clinics. In each case, when told of the errors, the ultrasonographer was surprised and dismayed that it had been impossible to even know that an error was occurring while performing this test.
An analysis of the clinical data showed that nearly 50% of the errors that we found in our experiments were a result of damaged probes. As long as the probe isn’t damaged so badly that it doesn’t work at all, then there is no way to tell by looking at them that there is a problem. Much of the rest of the error that we found was a result of improper actions on the part of the ultrasonographer. Each and every ultrasonographer that we pointed this out to was deeply concerned that they were unaware of the errors that they were capable of introducing. They were all very interested in further training and in the utilization of devices to practice on and test their equipment and themselves.
A properly designed quality assurance system in an ultrasound laboratory would allow the professionals running this laboratory to frequently check the accuracy of the processors as well as the probes. In addition, by having this equipment on hand, these facilities could give their professional ultrasonographers the hardware necessary to constantly retrain and test themselves so that self assurance and self certification could occur before any other type of evaluation or testing.