Transcription
Hello everybody, and welcome back. Let's finish off this dose and CT module by looking at the factors that influence patient dose. I said in the first two talks that it's actually quite difficult to calculate the specific absorbed dose that a patient receives during a CT examination, and as a result, it's almost impossible to confidently give a risk estimate to that patient for the ionizing radiation exposure that they receive during the CT exam.
Now, instead, what we can do is abide by what's known as the ALARA principle, and I'm sure you've heard of this before. Now, this is an incredibly important topic when it comes to CT physics because it has actual outcomes for the patient, and as a result, this question comes up almost in every single CT physics exam. And when I'm asked in an exam about the factors that influence patient dose, I think about dose in three separate domains: the factors that we can change prior to the patient receiving the scan, whilst the patient is still in their bed in the hospital. Then I think about the scanning parameters that we can change, and this obviously involves the most physics, and as a result, we're going to spend most of our time today looking at the scanning parameters. But I also think about the post-scan factors: what can we change after the patient has had their exam that will influence future patients' dose? Obviously, once the patient has received an exam, they've received that dose; we can't go and take away dose from that patient.
So let's start by focusing on the physics components here, the scanning parameters that influence patient dose, and there are five different parameters that I want to cover today, the five main parameters that influence dose. The first being what's known as the KVP, or the kilovolt peak. We know that when we place a patient within the CT scanner, we produce X-rays in the X-ray tube, and those X-rays head out towards the patient, and dose is distributed through that patient. We looked at that in the last talk. Specifically, we produce X-rays by accelerating electrons from the tungsten filament in the cathode towards the anode here. The number of electrons and the energy that they are accelerated between the cathode and the anode is determined by the KVP. Those electrons strike the anode and create X-rays in a spectrum. Those X-rays are consisting of Bremsstrahlung and characteristic X-rays, and that spectrum of X-rays contains a different number of X-ray photons at all the different energy levels. We've seen this diagram before where we have the electrons being accelerated towards the anode. This is a single tungsten atom within the anode, and those electrons are going to create X-rays here. These X-ray photons again create a spectrum, and you've seen that this block here, this purple block, is a filter. Here's the filter here that preferentially filters out the lower energy X-rays, so we get our X-ray spectrum heading towards the patient.
What happens then if we were to increase the KVP? Increasing the KVP is going to increase the number of electrons and the average energy of those electrons, and as a result, it's going to influence the X-ray spectrum. Now, importantly, dose is related to KVP. Dose is roughly proportional to KVP to the power of N. Now, why have I said to the power of N and not given you a number here? Well, in traditional radiography, we often say that dose is roughly proportional to KVP to the power of two. In CT scanners, N is normally closer to about 2.5, 2.6, depending on the filtration that we use, depending on the type of CT scanner that we use. But you can see that minor changes in KVP are going to have quite drastic changes in the dose to the patient, predominantly due to the number of X-ray photons that are heading towards the patient. Watch what happens here when I change the KVP from 80 kV to 100 kV. See how the X-ray spectrum has shifted here? The area under the curve has changed. Let me go back and watch how the graph changes from 100 to 80. Our graph gets much smaller. Our highest photon energy, our maximum photon energy, gets smaller. The area under the curve gets smaller. As a result, the dose gets lower. As we increase KVP, we increase the dose to the patient at roughly proportional to KVP to the power of 2.5 or 2.6. So KVP has a huge role to play when it comes to dose. You may remember the 15% rule when we looked at radiography, where we said that increasing the KVP by just 15% would require a halving of the milliampere of the filament current in order to keep the dose the same. A small 15% increase required a halving of the filament current. Here, it just shows you how much of a role KVP has to play in dose.
Next, we're going to look at filtration. We've seen that the filter preferentially removes the lower energy X-ray photons, and this is good because it not only increases the average energy of the photons that reach the patient, increasing the quality of the X-ray beam, but it also removes this dose that the patient would have received from these lower energy X-ray photons that would never have contributed to the image itself, because none of these photons would have made it through the patient. These low-energy X-ray photons would have been attenuated by the patient and never reach the detector, providing us no valuable information. When we remove the filter here, we can see that we are going to now expose the patient to much more X-ray photons. The area under the graph here, or the area under the curve, is much larger. And remember, we had different types of filters in CT scanners, the bow tie filter that compensates for the changes in patient shape. We know that as we head out to the peripheries, generally, of body parts, there's less tissue for the X-ray photons to travel through, and that shaping filter is going to allow for those changes in attenuation. But providing filtration is going to increase the X-ray beam quality as well as reduce patient dose.
Next, we're going to look at pitch, and we looked at pitch in the previous talk. We said pitch is related to the rate at which the patient moves through the CT scanner, the table speed, as well as how fast the X-ray source is rotating around the patient, as well as the beam width. If we were to keep the table speed the same and the rotation time the same, but just increase the beam width, increase the coverage in the Z-axis for one beam, increasing our cone angle here, reducing collimation, say we would be exposing more of the patient to ionizing radiation per rotation. So we're going to be reducing the pitch, and a reduction in pitch, we know, is an increase in dose. The same with table speed: the slower we move the patient through the scanner, the more they're going to be exposed to ionizing radiation for a specific coverage. The lower the pitch is going to be. So we can see that lowering the pitch is going to increase the dose. Dose is inversely proportional to pitch. The speed at which the patient moves through the CT scanner, the table speed, is going to determine how that dose is distributed in the Z-axis. So it's important to remember this dose-pitch relationship, but it's also important to remember how changing each one of these parameters is going to influence pitch and ultimately influence dose. A classic exam question again.
The next factor that we're going to look at is current modulation, milliampere modulation. We know that if we modulate the current that's passing through the tungsten filament within the cathode, we're going to change the number of electrons that are available at the surface of the filament through the process of thermionic emission, and ultimately change the number of electrons that are going to be accelerated towards the anode. That, in turn, is going to determine the number of X-ray photons heading out towards our patient. Now, if we were to plot a graph here of current on the Y-axis and time on the X-axis whilst the CT machine rotates around our patient, and we had a little sensor here by the detectors that sensed how many X-ray photons are being received by these detectors, if the patient had a constant diameter throughout, the number of X-ray photons reaching that center would remain constant. And if we were to link this sensor with our X-ray tube here, and we were to say, if the X-ray photon number changes here, we'll change our current, we could plot the graph that would be made for this specific exam. A patient of constant diameter requires a constant current at the cathode because the X-ray photons are being attenuated evenly, no matter what angle they incident on the patient.
Now, we know that patients aren't perfect circles. Generally, most parts of the body have an oval shape like this. X-rays passing through this angle of the patient are passing through a shorter distance, and therefore we require fewer X-rays to get the same exposure on the detectors compared to X-rays passing lengthways through this oval. We're going to need more X-rays passing through here to get the same exposure as a result. We're going to have to increase the filament current. Now, watch with this shape how the filament current needs to change in order to keep equal exposure on the X-ray detectors. You can see that as we pass through thicker parts of the patient, we require more current; thinner parts of the patient require less current. This is what's known as angular current modulation, current modulation based on the shape of the patient in the XY plane. That's important.
Now, if we place our patient in the scanner here, and we're rotating around them, and we've got an oval patient here, we're going to need angular modulation like this. But we know that firstly, our patient isn't always the same width. Their oval that they have is not the same width depending on where we are in the body. If we take our scout image here, or our scanogram here, the head is much narrower than say the thorax here. In this patient, their hips are even wider than the thorax. So we're going to need to change that angular modulation depending on where we are in the body. So, for instance, in this case, as the CT machine rotates around this patient, the tube current is going to have to change, or the filament current is going to have to change depending on the width of the different body parts here. And you can see those changes here again. This is what's known as angular current modulation.
Angular current modulation. If you think about a helical scanner here that's rotating as the patient's moving through the CT scanner, it's being exposed to different body parts. That's going to change based on the dimensions of the patient. Not only are the dimensions of the patient changing though, but the densities of the different structures based on location also changes. Notice how the thorax is much less dense in general than say the pelvis here. We're going to require a higher current here to get the same exposure going through the pelvis as we would going through the thorax. And what we can do is either place a sensor here that detects the number of X-ray photons traveling through the patient and is being received at the detectors, and ultimately feed back to our filament and increase or decrease the filament based on the amount of X-rays, or the number of X-rays passing through the patient. Or we could take a scanogram like this and analyze the density of the structures in this scout radiograph and say, well, let's take every column of pixels here and add up their pixel values, and that would correspond to the density of the various different structures based on their location in the patient. And that's exactly what I've done with this graph here that I've made. I've taken each of the pixel values and summed them for each of the columns in this scanogram, and I've plotted that on a graph. Notice how the petrous bone has a little bit more attenuation than say the skull here. The neck has very little attenuation. The shoulder girdle, because we've got bones, we've got the clavicle, the scapula, the head of the humerus here, it's going to be a bit more dense. Lung tissue, less dense. Breast tissue is contributing to some density here. We get a drop-off below the breast tissue before we then get into the solid organs of the liver here, and we see the pelvis has even more attenuation in this patient. This is what's known as modulation in the Z-axis.
So we need to modulate tube current or filament current based on the dimensions of the patient here, as well as the density of the different structures in the patient here. This is analogous to automatic exposure control that you may have seen in radiography. Here, this is what's known as tube modulation or current modulation, and we can combine these two factors to get our overall current modulation that we're going to need for this specific patient. Notice how we still get that oscillating sine wave here based on the dimensions of the patient, but our baseline current is going to change based on the density in the Z-axis here. This is perhaps the most difficult physics concept to understand in the factors that influence dose, and I'd encourage you to spend some time figuring out in your mind the difference between the angular modulation and the Z-axis modulation.
Now, another factor that may come up when we're talking about dose reduction and CT scanning is what's known as shielding. There are certain areas in the body that we know are more radiosensitive, like gonads, because those genetic mutations can be passed on generations, or say breast tissue is more sensitive to dose, and we can place shields over those tissues when the patient is going into this imaging machine and place a bismuth shield, say, over the breast tissue. And that, in theory, if we were taking an abdominal scan, would prevent some X-rays from being incident on the breast tissue if we were to creep up into the chest cavity for that CT scan. So that reduces dose from incident X-rays; it doesn't reduce dose from scattered X-rays within the patient. Those scattered X-rays are going to go underneath that shield that we placed across the patient. The reason I'm mentioning it here is often you can make the mistake of placing the shield on the patient when you take the scout image. That shield is very dense; it doesn't allow X-ray photons to pass through it. Then, when we run this current modulation, we're going to include that X-ray shield in our calculations for this Z-axis tube modulation. That's going to ramp up the filament current way higher to try and penetrate that shield and inadvertently is going to expose our patient to more dose. So placing the shield can, if we're not careful, end up being more harmful to the patient than it is good. We must remove the shield when taking the scout image if the scout image is what's being used to modulate the filament current. That's an important point to remember. Obviously, shielding is incredibly important when it comes to staff members who are taking the CT scan or taking X-rays. That's something we're going to talk about when we talk about radiation safety later on.
Now, the last concept I want to cover when it comes to scanning parameters is a very simple one: it's coverage. Obviously, the scan length that we expose the patient to is going to be proportional to the dose that that patient receives. The more scan length, the more coverage we have of the anatomy in the patient, the more dose that patient's going to receive. This is something that can easily happen in clinical practice. Say we want to get a scan of the patient's liver, and you think, I'm just going to include a little bit more in that scan, I'm just going to include a little bit more, and eventually, when you want to look at, say, just the abdomen, you start looking at pelvic structures, you start looking at the lower thorax here, just to make sure that we include everything. We're not doing the patient any service because we are now exposing them to more radiation than their specific examination requires. So this is an important factor to remember when planning a CT scan.
So those are the scanning parameters. I said there were some pre-scan factors and post-scan factors. These don't really have physics principles, but they're important to remember when thinking about the ALARA principle, when thinking about reducing total patient dose. Firstly, before the patient receives a scan, we need to say, is the scan justified? Will performing this scan change an outcome for the patient? Will it change the management for that patient? Does this patient actually need this specific scan? That's important to remember. We're very quick to say every patient coming into the ED just needs a CT scan. Next, is there an alternative investigation? Is this an at-risk patient? Is this a young patient? Are there alternative investigations that don't use ionizing radiation that we have at our hospital and are feasible for us? Next, we need to think about the type of scan that we're going to do. Are we going to have a three-phase scan? Are we going to have a pre-contrast, a contrast, and a post-contrast scan, or can we get away with, say, an uncontrasted scan? Obviously, increasing the number of phases repeats the number of times we're doing the CT scan itself, and every repetition is just adding dose to that patient. And lastly, and perhaps most importantly, is to check for priors. This is something that we as radiologists, perhaps, are most guilty of, or ordering clinicians are most guilty of, is you have a patient coming in for the same problem week on week on week, and they're getting a CT scan every single week. That's coming back negative, and we're expecting a different result to come up. If this patient has multiple negative CT scans, perhaps they don't need the CT scan. Say, a pregnant patient that's come in with shortness of breath, and we've checked for a PE multiple times. Do they really need another CT scan, or do we need to start looking for another cause for that shortness of breath?
And then I think about post-scan factors. These things don't influence the patient that's had the CT scan right now's dose. Obviously, if you've had the CT scan, we can't go and take away dose; you've received the dose that you've received. But it's going to affect future patients that are receiving CT scans in our unit. The first thing we can look at is the hardware that we have within our units. Are we using the appropriate filtration? Is this an old CT scanner? Is it time for us to upgrade to newer, better equipment that perhaps doesn't expose the patient to as much radiation? Are there defects in the hardware? Are all our detectors working? Are we having to compensate for some faults in the hardware that means that we require higher tube currents to get the same image quality? Next, we can think about the software. We know that say, for image reconstruction, we saw that iterative reconstruction techniques can allow us to perform CT examinations at much lower doses but still get the same image quality. So perhaps updating the software is going to reduce the amount of dose needed to create the same image quality. And lastly, this is something that's important, especially for new radiologists who are wanting to do research in their unit or want to make an impact on patients within their unit, is to review the quality of the images. Perhaps have ordered protocols in place. So, of all of the images that we've taken so far, are they of adequate quality? Are we able to reduce certain parameters and still get the outcomes that we need for the patient? Constantly reviewing, are we applying the ALARA principle to our patients? Are we giving them doses that are low as low as reasonably achievable while still maintaining diagnostic quality?
So that brings us to the end of this dose in CT module. We've now looked at three separate talks, and hopefully you understand what dose is and how it occurs, how we go about calculating dose and estimating risk, and the factors that we can change in order to reduce patient dose. Now, we're going to shift gears entirely and look at a type of CT acquisition known as dual-energy CT. So I'll see you all in that talk. Until then, goodbye everybody.