Transcription
Let's take a closer look today at spectral Doppler ultrasound. We've looked at the concept of the Doppler effect and Doppler shift within tissues, and we've seen how we can use that Doppler shift value in order to measure the magnitude of movement within tissues. And if we know the angle of our ultrasound pulse versus the angle of the movement within tissues, we can use that Doppler angle in order to measure a specific velocity within tissues.
Now, spectral Doppler is the display of velocity change over time within a region of interest. Now, in our previous talk, we looked at continuous wave and pulse wave Doppler ultrasound imaging, both of which can create spectral Doppler ultrasound. Now, today I want to first cover how we go about creating a spectral waveform, and then touch briefly on some measurements that we can do within a spectral waveform, and end things off with two very brief clinical examples.
Now, the focus here is going to be on the physics. We're not going to be focusing on actual clinical outcomes, and all of my schematics here are diagrammatic representations of the spectral Doppler waveform. Nothing here is extremely accurate. I want you to get the concepts over actually measuring specific values.
Now, in order to create a spectral Doppler waveform, the first thing we generally do is create a B-mode image, a brightness mode image with a standard ultrasound transducer. That allows us to appreciate the underlying anatomy within the patient's tissues. Then, often we place what is known as an active area over our region of interest, and this is a color Doppler zone, a zone where we can see movement within the tissues. This often helps us to identify our vessel of interest and helps us to place our Doppler gate within a very specific region in that vessel.
The next thing we do is choose an A-line of data, a single A-line within this B-mode image that we are going to use pulse wave Doppler in this example to create our spectral waveform. Once we've chosen that A-line of data, we can then place what is known as a gate. Now, the gate specifies two things: it specifies the depth in the tissue that we want to image, and it specifies the volume of movement we want to image. The distance between these two gates here will be the area that we're sampling for the Doppler shift.
Now, once we've placed the gate, we need to tell the machine what angle our A-line is in comparison to our vessel of interest. And this setting of the angle means that we know what our Doppler angle is in the image. Now, this line, this angle calculation needs to be parallel with the blood flow within a vessel. So now we've set up our pulse wave Doppler image, we can go about creating these spectral waveforms that will form from blood moving between this gate here.
Now, the spectral waveform is made up of two axes. The first axis is time. The spectral waveform is the change in velocity over a period of time, and on our machine, we can change the time scale, known as the sweep speed, in which we display that spectral waveform. Now, in this example, say this was representing two seconds of data, we could see that this patient's heart rate is 120 beats per minute. Now, we could change this to display four seconds of data, or to display one second of data, and that will stretch and compress our spectral waveform accordingly.
Now, the Y-axis represents the velocity of blood within the spectral gate. We can see that the Doppler shift that we measure within that gate is proportional to the velocity of blood moving through that gate, and we'll see later how we can change this velocity scale by changing the pulse repetition frequency of our pulse wave Doppler.
The next component of the spectral waveform is the baseline, the level which represents no flow of blood, zero velocity. Now, the baseline can be moved on our screen, the screen that's displaying the spectral waveform. We can move that baseline in order to better see the spectral waveform. If all of our spectral waveform has positive velocities, we can move that baseline down slightly to see the spectral waveform better, because at some point we're going to be measuring specific points on the spectral waveform in order to calculate specific indices that have a clinical value.
Now, another thing that the spectral waveform displays is direction of blood flow. By convention, we say that a positive velocity represents blood flowing towards the transducer, and a negative velocity, blood flowing away from the transducer. Now, on a machine, we can actually change the direction if we want to. It's not changing the direction of blood flow, it's changing the way we display that blood flow on the spectral waveform.
Now, the actual spectral waveform, the line itself, can be broken down into various different measurements, and I've labeled the most important ones here. Now, when we think about how this line is created, there are multiple red blood cells, thousands of red blood cells passing through this gate at any given period of time. Our incident ultrasound beam will be reflecting off all of these thousands of red blood cells. They will each be sending back different reflections back to our ultrasound transducer. Now, all of those reflections heading back will interfere with one another. We'll be getting one continuous stream of data heading back towards our ultrasound transducer, and that stream of data will be a combination of all of the hundreds and thousands of waveforms heading back from the various different red blood cells. And we can use a process known as fast Fourier transform to take that one continuous data stream, and it will break it down into multiple different individual waveforms heading back. Now, those individual waveforms will represent specific Doppler shifts based on the red blood cells that have reflected that ultrasound wave back.
Now, we've got a range of velocities. Each red blood cell will have slightly different velocities. Now, the thickness of our spectral waveform line represents the differences in those velocities within the red blood cells at that given point in time. So you can see that if there was turbulent blood flow through the gate here, we would have a very thick spectral line. There will be a wide range of velocities at any given point in time. Some red blood cells actually may be flowing in the opposite direction, while some might be flowing quickly towards the ultrasound transducer, and this spectral waveform will spread over a wider region. The narrower, or the thinner, our spectral waveform line, the more uniform the velocity of those red blood cells heading through the spectral gate here.
In terms of labeling the various components of the spectral waveform, the first thing we need to know is the peak systolic velocity, the highest velocity within our blood flow at any given period of time. Then we look at what's known as the end-diastolic velocity, the velocity of blood flowing through our gates at the end of diastole, just before the next systole. Now, this is an arterial waveform that we're looking at, and the waveform shapes will vary depending on the type of vessel that we're imaging, and that's out of the scope of this talk.
Now, the window below this waveform during systole is what's known as the spectral window. And there are various reasons why that spectral window might be obliterated, might be filled in by signal. One of which, as we've mentioned, is turbulent flow. We can also increase the gain on our machine too high, where we start getting noise back and filling this spectral window falsely here.
Now, what we're representing with the spectral waveform is the change in velocity over time. Now, what is the change in velocity over time? That's the acceleration of an object. Now, we can see this line here, which is known as our systolic upstroke, and this represents the acceleration of blood during systole. Now, we can see that after we've had that contraction, that systolic contraction, and the heart relaxes, now that blood then decelerates. The velocity changes over time. So any movement in the spectral waveform that represents increasing velocities shows acceleration of blood, and the opposite is also true: decreasing velocities shows deceleration of blood.
Now, importantly, blood that is decelerating doesn't mean it's reversing here. It's slowing down. If we're in a car and we brake, we start to slow down, we start to decelerate, but the car is still moving forward. This blood on the downstroke here is still moving forward, and this spectral waveform actually represents blood that is continually moving forward, moving towards our ultrasound transducer. Even at the end of diastole, there is still a positive velocity here.
The last component that I want to mention is what is known as time to peak, the period of time it takes from the beginning of systole to get to our peak systolic velocity. Now, that time to peak is related to the acceleration of blood, and the faster the acceleration of blood, the shorter that time to peak is.
Now, importantly, this spectral waveform that we are creating here is specific to the gate that we have set, and we can change the size of that gate. Now, we know that blood flowing through a vessel exhibits a specific type of flow. Blood flowing in the center of that vessel will exhibit what's known as laminar flow. It will have the most consistent flow at the highest velocities, and the more peripheral we go out on that vessel, the slower the velocities and the less consistent those velocities are. Now, if we were to increase our gate size here, we can see that our spectral wave thickness would increase. The range in velocities that we detect here will increase. Our peak systolic velocity won't change, and our end-diastolic velocity won't change, but the thickness of that spectral waveform has changed because we are sampling a greater range of frequencies in a wider gate.
I mentioned earlier that the velocity scale here is determined by our pulse repetition frequency. When we change the velocity scale, the look of our waveform would change. Here, we've only changed the scale here. The maximum velocity that we are now displaying is 250 cm/s, where previously it was 120 cm/s. Now, nothing has changed in the velocity of the blood here, only our scale has changed.
Now, we are going to use the peak systolic velocity and the end-diastolic velocity to calculate what is known as resistive index. And we need to be really accurate when setting those measurements, and using a scale that is incorrect for our waveform means we're going to be less accurate here. We want our waveform to take up the entire real estate here so we can be accurate with those measurements.
Now, how do we go about actually calculating the scale here? When we set our A-line and set the gate here, what we are doing is inadvertently setting our pulse repetition period, the period of time between the first pulse and the next pulse. We've set a specific depth here, and that depth will allow us to determine how much time we have to wait in this receive time in order for the pulse to come back from our gate, the round trip distance here. So, say, for example, this was 4 cm deep in the tissue, we can then calculate how long it would take our pulse to go 4 cm in and return 4 cm back before then releasing the next pulse into the tissue. That's what's known as our pulse repetition period. Now, here our pulse repetition period has been set because of the depth that we are sampling within tissues.
Now, the pulse repetition period is inversely proportional to the pulse repetition frequency, the number of pulses sent into a tissue per second. Now, the higher our pulse repetition frequency, the shorter the pulse repetition period when we are sampling shallower tissues. Now, as that pulse repetition frequency increases, we can detect higher and higher Doppler shift values. We looked at in the previous talk that the Doppler shift value, the maximum Doppler shift value that we can detect is half of our pulse repetition frequency, and that Doppler shift value corresponds to a velocity value within tissues. So the higher our pulse repetition frequency, the higher the maximum velocity that we can sample.
Now, if we want to change this velocity scale in order to change it back to the maximum velocity being 120 cm/s and getting the waveform filling the entire screen here, allowing us to more accurately measure our peak systolic and end-diastolic velocities, what we need to do is lower our pulse repetition frequency. And how do we lower the pulse repetition frequency? We increase our receive time here, we increase the pulse repetition period. Now, we're sampling that tissue at a lower rate, our pulse repetition frequency has decreased, but because the Doppler shift in this blood at the set velocity is still less than half the pulse repetition frequency, we can still accurately map those velocities over time. So changing our pulse repetition frequency will change the scale that we use, the velocity scale on our Y-axis in spectral Doppler.
I've said earlier that we will use these values to calculate what is known as a resistive index. Now, the resistive index takes the difference between our peak systolic velocity and our end-diastolic velocity and divides that by the peak systolic velocity. Now, we can see here that if the end-diastolic velocity is a positive value, our resistive index will always be less than one. If the end-diastolic velocity ends up being zero, if there's no flow at the end of diastole, our equation here will be our peak systolic velocity over our peak systolic velocity, our resistive index will be one. If there's reverse flow at the end of diastole, i.e., the end-diastolic velocity is a negative value, the numerator here will be more than our denominator, our resistive index will be more than one. Our peak systolic velocity subtracting a negative value is the same as adding that value. We've got a larger numerator and a smaller denominator.
Now, the resistive index is a proxy for downstream resistance. We think of blood flowing through a vessel as the proximal portion of that vessel being upstream. As we head through that vessel, we think of that as downstream. The resistive index shows us how much resistance there is downstream. So we can see that a low resistive index means low peripheral vascular resistance, and the higher the resistive index, the higher the peripheral vascular resistance.
Now, we use this resistance. Our body uses this resistance in order to titrate or shunt blood to where it is needed the most. Now, there are certain organs within the body that need a constant flow of blood, they're high metabolic demand organs. We know these organs such as the brain, it needs continuous blood flow. The kidneys, the liver, they need a low peripheral vascular resistance. In normal blood flow, we will always have a positive end-diastolic value. There is continual forward motion of blood to these high metabolic demand organs.
Now, there are certain tissues like our limbs or the bowel that doesn't need continual flow of blood. We can have higher peripheral vascular resistance in these regions so that we're not shunting blood towards our muscles when we're at rest. Now, this can change if we start to exercise. We start to run, the capillary beds, the vascular beds within our muscles are going to dilate, they're going to reduce that peripheral vascular resistance, and our resistive index will decrease. We will shunt blood to where it's needed. So our body can change the resistance of those peripheral vascular beds in order to move blood around the body. We know that blood moves from higher pressure regions to lower pressure regions, and that's a really important concept to remember.
Let's go ahead and look at a specific example. If we take the common carotid artery here, that then splits into the external carotid and internal carotid artery. Our internal carotid artery is going towards the brain. It needs to supply a constant blood flow towards the brain. Our resistive index here is going to be less than one. Our peak systolic velocity and end-diastolic velocities are both positive values. There's continual forward motion of blood. The velocity of this blood never crosses the baseline.
Now, our external carotid artery has multiple branches, many to superficial structures on the head and neck. Now, during diastole, at the end of diastole, there is no flow within those vessels. Only during systole and the first part of diastole is there forward flow within the external carotid artery. If the end-diastolic velocity value is zero, our resistive index will be one. If there's slight reverse flow, our resistive index will be more than one.
Now, what happens when we're scanning the internal carotid artery looking for atherosclerotic disease? We will see changes within the spectral waveform that will help us to realize whether there is atherosclerosis within that vessel or not. We're looking at if there's stenosis within a vessel, and this is one of the most common applications of spectral ultrasound. So let's zoom in on an internal carotid artery here that has a thrombus within the wall. Now, we can see that the blood flow heading in the proximal section, the upstream section of our internal carotid artery has slightly changed here. We've increased that peripheral vascular resistance. We've seen that as we increase the peripheral vascular resistance, we increase the resistive index of the spectral waveform. So now we are still getting this peak systolic velocity coming from the heart here, but in diastole, that end-diastolic velocity has now reduced. It's coming towards zero here. This resistance from the stenosis is causing that diastolic forward flow of blood to slow down. We are reaching the baseline here.
If we were to place a spectral gate right in the middle of this stenosis here, we would see the velocity values that we calculated are way higher than this upstream or proximal section. If you have a hose pipe that's got water coming out of it, and you occlude some of that hose pipe with your thumb, you're not changing the velocity of water coming to your thumb. That velocity suddenly becomes much faster as you narrow the lumen, or as you narrow the radius of where the blood is flowing through, or where the water on our hose pipe is flowing through. Now, what you'll see here is that the velocity we measure is beyond the scale that we've set on our machine, and we get a phenomenon known as aliasing here, an ultrasound artifact that's specific to Doppler. In our next talk, we're going to be looking specifically at aliasing and how we can go about reducing that aliasing artifact.
The blood that makes it through the stenosis will now be turbulent. It's going from a narrow lumen to a wide lumen, and it's going around the thrombus. So that blood starts to become turbulent, and we see now that we've reduced the peak systolic velocity and we've closed the spectral window. There is turbulence within that blood. There is a larger range of velocities within the blood heading out of the stenosis.
Further downstream, we get a very characteristic waveform that is a downstream effect of a proximal stenosis, and this is what's known as tardus parvus. Now, what we have here is a lower peak systolic velocity. We are getting reduced velocities downstream from the stenosis. We also have a slower acceleration time here because only a set amount of blood can make it through this tight stenosis. So this slower acceleration to our peak systolic velocity and a higher end-diastolic velocity means the resistive index here has been decreased. Now, why is this diastolic velocity so high? Well, the brain has been starved of oxygen here. We are not getting enough blood supply through this internal carotid artery, and as a compensatory mechanism, we've reduced that peripheral vascular resistance, and that reduction in resistance, that reduction in vascular pressures means that we get forward flow in the low pressure diastolic state. We are trying to increase that blood flow to the brain. So we can see how a pathology leads to spectral waveform changes, and how we can look at these spectral waveforms and try and explain them with underlying physics principles.
Let's look at one more example. This is the portal vein heading towards the liver. It's taking blood from the gut and from the pancreas and from the spleen and taking it towards the liver. Now, in a normal spectral waveform here, we have a fairly thick spectral waveform, multiple velocities heading towards the liver, and they're generally at lower velocities, 15 to 30 cm/s. Now, that waveform has a slight phase to it, and that phase is dependent on both the cardiac cycle as well as our respiratory rate. As we breathe in and out, we are changing intrathoracic pressures, and that also changes our venous pressures. So we can see the slight phase to the waveform.
Now, what happens if we get something like portal hypertension? There's increased resistance to that blood flowing back to the liver. We get a reduction in the velocity of blood within the portal vein. We can see now that those velocities have decreased. If there's no velocity within the portal vein, we need to be worried about something like a portal vein thrombosis. We've actually got an occlusion and no flow of blood within the portal vein. We can also get an increase in the variation within the blood flow in the portal vein, and this occurs in something like tricuspid regurgitation, where the heart goes through a systolic contraction, and there's actually reverse flow of blood through that incompetent tricuspid valve down into our venous system, into the inferior vena cava, and into the liver. And we can see that there's some reversal of flow with those increased pressures within our inferior vena cava. So during systole, there's reversal of flow, and during diastole, we are getting that normal flow that we were seeing in our initial waveform here.
Now, this talk is not meant to teach you the different types of waveforms in the different types of vessels. I just want to show you how we can use the basic principles that underlie spectral Doppler and apply them to clinical scenarios. And in the future, we're going to go through multiple different clinical scenarios and multiple different vessels throughout the body, but that is for another time. And if you're studying for an ultrasound physics exam, focus on the initial part of this talk: how we go about creating that spectral waveform and which parameter changes will lead to changes within the spectral waveform. And in the next talk, we're going to look at the concept of aliasing, where our pulse repetition frequency is not sufficient for the velocities within the blood, and we need to make different changes in order to compensate for that aliasing.
Now, the concept of spectral waveforms and aliasing comes up over and over again in exams, and I've linked below a curated question bank for those of you that are studying for an exam. So go and check that out if that is you. Otherwise, I'll see you in the next talk, where we're going to take a deep dive into what aliasing is and how we can go about reducing aliasing within the spectral waveform. Until then, goodbye everybody.