Doppler Features,Pitfalls, and Artifacts - HD

Doppler Equation

We'll start this off.

What I'm gonna talk about here is a little bit about the Doppler equation, because this is central to understanding the Doppler work that we do.

I would argue that if there's any one equation in all of radiology that you might wanna remember, this is the one, because this actually has a lot of practical utility.

What this tells you is how to tweak your machine if you're having trouble getting Doppler signal.

The sensitivity for the Doppler shift is directly proportional to the fundamental frequency.

If you go up in frequency, you should be able to detect slower flow, and it's directly proportional to the cosine theta.

To remind you that the cosign of zero is one, and the cosign of 90 is zero.

If you want that maximal Doppler experience, you want to be at zero degrees incidence, you wanna be on the train tracks while the train is coming at you, you'll have the maximal Doppler experience before it hits you, and it's all over this speed of sound, which is a constant.

Of all those things that you might remember, this one is worth remembering.

To remind you, the angle theta is the angle of incidence to the blood flow.

The f on the right side of that equation has to do with the fundamental frequency of the transducer.

A consequence of that is that the Doppler shift increases with increased incident frequency, increased reflector speed, which is of course what you're trying to measure and decrease incident angle.

You probably all know by now that the optimal angle is somewhat less than 60, because above 60, the variability of the measurement really becomes quite high.

Pitfalls in Doppler

Let's talk about some of the pitfalls, and these I've selected really to talk about how doppler can be improved if you're having trouble.

There is an inherent variability of doppler measurements.

Most of it's related to technique, I suppose, but there are instrument variabilities.

People have looked at different machines and shown that you can get slightly different values with different machines.

There are certainly physiologic differences in patients.

Vessel and plaque morphology can contribute to some of the variability that you see.

Things as simple as gain setting can change what you will see record and measure.

Gain Settings

Here's an example of that.

This is a gain setting that was changed from this one here to something much less.

You can see that if you took these at face value, and obviously this one is starting to get a little bit on the weak side, you could get a measurement difference of 15 or more centimeters per second.

Optimal gain settings has consequences in spectral doppler and those measurements, but also within color doppler.

Here I've changed three things.

Again, this is I did this in perhaps in an exaggerated way to make a point, but you can see that if you under gain, you will miss some of the flow around the wall.

If you over gain, you can bleed out into the adjacent tissues and this would be more optimally gained.

There are real consequences to all that.

Here. Again, gain settings at three different settings, and you can see how radically that can change.

If you have a new sonographer and they're not getting flow, say in the portal vein, one of the things to be alert to is have they optimally gained, is the gain setting appropriate?

The way that I tend to do that is to turn up the color gain until you get that those flurries like this down here and then back off until you no longer have those.

Another consequence of that would be, here's this in B mode.

You can see that there's a lot of echoes in this, and the plaque here is somewhat obscured within this background of echoes.

Here this is the color doppler, which I have turned too high, and I've actually obscured a good portion of that plaque.

It does have consequences.

Here's another color gain settings that are changed.

You can see that the peak systolic velocity is measured in this sample here somewhere around 54.

When you turn up the gain, you can get a change of to 129.

I think that makes the point that gain settings are important and can be a source of interpretive error.

Center Stream Sampling

Center stream sampling.

We know when we sample in the carotid artery, it's important to sample in the center stream, not too close to the wall, because then you'll pick up a lot of wonky wall movement.

Here is the here I put the doppler cursor right against the wall, which of course you wouldn't do if you were paying attention.

You can see you start to get really unusual waveforms when you place the cursor or the range gate rather closer to the wall.

You pick up that wall motion, whereas in center stream, you get something quite different so that even the appearance of the waveform changes.

You can if you turn up the if you open the range gate too high, and this again is sort of ridiculously high, but it makes the point.

I've opened the range gate to the to encompass the entire carotid artery.

You get something that shows, well, I guess you could argue spectral doppler spectral broadening and an unusual waveform.

'cause you're picking up that wall motion.

Wall Filter

The way that the machine handles wall motion is there is a wall filter, and you can change that too, so you can filter too much out if you are so inclined here is the wall filter here.

It which you'll see it it represent it'll represent you'll see it as this band of blackness at the baseline.

Here I've turned it up a little farther and here even farther yet, and you can start to see how you can really remove essential parts of the waveform if the wall filter is set too high.

Doppler Angle

Anybody who does doppler knows of the importance of setting the Doppler angle correctly, and how much that can change things.

You may not be aware of just the magnitude of the change that even a relatively minor doppler angle difference can be.

Here's the Doppler angle appropriately set.

I'm getting a measurement at 144 centimeters per second here.

This is really not far off, it's not right, but it's not so terribly far off, and you can see that it's nearly 200 centimeters per second.

Even relatively small mal adjustments of the Doppler angle can lead to relatively large changes in the measured velocity.

Here's that here's a time honored graph to how the the accuracy of the measured waveform degrades above 60 degrees, and it does so rather rapidly.

One thing I might point out with this is if you are at 90 degrees, you often get these very strange triangular waveforms or diamond shaped waveforms where you get representation above and below the baseline.

I think that's caused by probably one of two mechanisms.

One, the machine doesn't know whether the blood flow is going too towards or away from it.

So it will represent it on both sides.

Secondly, remember that there is spread of the beam and on both sides, so it is picking up some elements that are coming towards it, some blood elements that are leaving away from it.

This sort of diamond shape, if you get that waveform, you've got an angle problem.

Keep that in mind.

Here's a couple more examples of how important angle is.

This is a measurement of the carotid without any angle adjustment.

You can see this is over a thousand centimeters per second, which would certainly get anyone's attention.

But if you put the angle correction on it is something closer to a hundred, which is much more reasonable.

You can see how that how why a lot of Doppler experts are militant about the Doppler angle because if ever they see a study submitted for publication and the Doppler angle is not appropriately adjusted, then basically it's garbage.

Here it is.

The portal vein is notorious in this regard because it is almost always a 90 or close to 90 degrees incidence to your transducer.

Slow flow in the portal vein is a notorious example of where the Doppler angle can lead to considerable misinterpretation.

Aliasing

A sine this is one we all kind of know about, it's basically an under sampling error.

Least occurs when the pulse repetition frequency, which is the sampling rate, is less than two times the doppler shift.

Here it is diagrammatically.

Here's the change of the doppler shift that you're measuring.

If you under sample it, you're going to represent it as something that's changing much more slowly than it in fact is.

What you might not be aware of is that every time you look at a Doppler thing, you are actually looking right at this nyquist limit.

That's the limit at which lysine occurs.

Here it is on the on the spectral doppler.

It's right here in this scale.

In color Doppler, it's on the color bar.

The point being is that when you change your velocity scale, what you really are doing is changing your PRF.

Here's an example of that.

When they used to display this, here's the velocity scale, it's up to 60 in the forward flow direction, and the PRF is 1221.

What we've done is chopped off this the peak here, and it's written below the baseline here, but instead of being downward facing it it what has happened is it is wrapped around, that's the alist part.

If you change your velocity scale, which I've done here, suddenly, I don't no longer get that truncation of the of the wave form, but rather I have it fully upright.

Look at the pulse repetition frequency, it's jacked up to over 2000.

What you're doing when you adjust that velocity scale, it's not just changing a scale, but you're actually changing your pulse repetition frequency.

How do you correct for aine?

There's a couple of ways you can change the baseline, which assigns more of that scale to the forward flow.

You can change the velocity scale, in which case you're changing the PRF, or you could use a lower Doppler frequency or adjust your angle to a higher angle.

Just don't go above 60.

Lysine is not limited to spectral doppler.

Of course, it happens. And color as well.

The way you know that you're looking at lysine is that when you do get that color change is going through a white line, not a black one, the black one is true flow reversal.

A white line indicates sine.

Spectral Broadening

That brings the case of spectral broadening.

Remember that most of the flow normally in larger vessels anyway, and the body, has a relatively laminar configuration.

This is obviously a vein, but it's cool.

Anyway, I was gonna show you, but that you get some sense of how that flow pri profile looks, you know, laminar there.

Spectral broadening occurs in areas where there is non laminar flow or disturbed flow.

It's seen in areas where there are bifurcations, where of course, where there's disease, and it's also seen in smaller vessels.

This spectral this spectral broadening that we talk about is represented in the spectral doppler waveform as this filling in of velocities underneath the envelope of peak systolic velocities.

You get that ac something sometimes called the acoustic window filling in.

Where can you get the evidence of spectral broadening that's not really representative of non laminar flow, but is a pitfall?

Of course, we've already seen gain settings, high gain settings can do it.

If you pick up wall motion, if you're if you're not in the center stream, you can get a waveform that demonstrates spectral broadening or when the vessel abruptly changes caliber or direction.

Laminar flow is always distorted at bifurcation.

It could come here.

As it wraps around a tight curve, you'll get the laminar flow right in that area.

You can already see that the representation is non parabolic, and in fact, the velocity in the outer portion of the vessel will be higher if you measure it than on the inner portion.

That's just good old physics.

The consequence of it is that measuring around an abrupt curvature is problematic for our machines.

As we all know, some of the vessels are extraordinarily tortuous as they are in this case.

It gets very hard to know how to adjust your doppler angle in a vessel that tortuous.

Helical Flow

One phenomenon that you may not be aware of, you might have seen, is something called helical flow.

This happens around areas of bifurcation, sometimes areas of disease.

It's a really cool phenomenon.

What you'll see is if you're 90 degrees on it in transverse, you'll get a red and a blue.

You'll get two different colors because this is flow that is spiraling, it is helical, like a barber pole.

If you angle your transducer radically up or back, you'll get a solid color because there's two vectors of movement here.

There's the vector that of course is down longitudinally, the length of the vessel, but there's also the part that's spinning.

There's part of it that's coming at you and part of it that's going away from you in a true transverse.

What's really interesting about this is that if you go from side to side, you can get that color change from red to blue and red to blue, even in longitudinal, because it is in fact spiraling or spinning flow.

What you want to be sure of is that that's not a dissection, and it could be really confusing, because here's a dissection here, there's a little thin band.

In a similar way, you get different flow profiles, but it's not simply a change of direction.

Usually in the setting of a di of a a dissection the flow profiles will look radically different from side to side.

Here's an example of it here in the false lumen, there's no way that that's a normal waveform in the ICA.

It's worth looking at the character of the waveform because that will help you distinguish helical flow, which is normal in some cases.

And dissection, which is of course abnormal.

Boundary Layer Separation

Closely related to helical flow is something called boundary layer separation.

What that means in the world of hemodynamics is that when you get areas of laminar flow, but then it you have an abrupt change either in direction or behind some sort of obstacle, you can get eddy currents behind it, and the slower boundary layer can in fact reverse.

Here it is.

Here's the parabolic or laminar flow that we spoke of.

It's slower, close to the wall.

It's faster in the center stream.

Any point where this bulk flow is gonna slow, like say it reaches the capacious ICA bulb, then the all the elements of that flow are gonna slow down.

In that portion of the flow, the boundary layer can in fact reverse.

This was discovered or described fairly early on in the world of doppler, but you get up to the bulb and if you sample at the edge of the bulb, you'll get true flow reversal.

That's not aliasing because look, it's going through a black line.

That's true flow reversal that you can see at the edges of capacious or widening areas or at the behind disease.

Boundary layer separation is a real phenomenon.

You will see it don't sample, you will not get meaningful results in that area.

Here's an example of a very similar sort of ed current around an ulceration.

Measurement Abnormalities

A couple more things to say about pitfalls.

That one would be abner abnormalities that come from measurements.

Not the same.

The measurement of the luminal diameter will not be the same in gray scale power and colored doppler.

That's because of the very of what they're doing at that edge of the vessel.

I'll get into that in a minute.

We don't tend to measure diameters of stenosis, but there has been some enthusiasm for what's something called the minimal luminal, a minimal residual lumen, or measuring the how much there is left.

If I remember correctly, if it was 0.6 or below that indicated critical stenosis.

Nonetheless, there is some people, some people measure the luminal stenosis itself, reminding you that a lot of the criteria, the NASA and the ACEs and all those sort of things that Leslie Scout talked about yesterday, really derived from angiographic measurements.

They measure diameters of the patent lumen to relative to something downstream, a normal downstream lumen.

With ultrasound, of course, we can see the true narrowing, we can see the walls on both sides, and we can obtain a true measure of the narrowing at that one point.

If you're a fan of that sort of thing, you have to be a little bit careful one.

We have already alluded to is that's the wall filter.

If you set that too high, you will narrow your estimate of the patent lumen.

But also, if you choose color or power or gray scale, you'll get different numbers.

Here's an example of that. This is color doppler.

This is power doppler.

You can see that the color doppler looks wider.

You would measure this would be wider than if you measured it in power.

Which one is the more accurate?

The answer is the power is more accurate.

The reason for that is that power doppler, which measures the intensity of the returning signal is what's known as continuous boundary discriminator.

It can show you it the intensity of the representation decreases this you approach the boundary.

Color, however, is what is known as by stable.

What's happening here is like here here it is tally, it's the mean velocity within that pixel that's represented.

That pixel of it overwrites the wall, and some of them will be all written the same color.

As a result, you'll get larger diameters or measurable diameters with color doppler than you will get with power or gray scale.

That is statistically significant.

This is a study we did some time ago, and it's not a lot, but it is certainly real.

If you have no do doppler signal detected, what can you do?

Either your scale's too high, your gain's too low, either your frequency's too high or low, depending on whether you've you've either shrunk it so far that it's not visible or made it so big that you can't see the full extent of it.

Power to low wall filter too high.

Artifacts in Color and Power Doppler

Now I'm gonna talk a little bit about artifacts and end with that.

These are artifacts specific to color and power doppler.

This one is not, this is mere image artifact.

We've all seen that in gray scale, but I just want to point out that there are mere image artifacts that occur in color doppler, and here it is.

Here's a duplicated ECA because of the high reflectivity of the calcified wall.

This is an iliac artery that was duplicated.

What's interesting about these things is that the color is mostly replicated the in the mirror image as it is.

What's required of this is that there'd be a highly reflective interface somewhere around the vessel.

The subclavian artery is a good candidate for that because it arches over the pleural surface.

You can get it written into the lung itself.

Refractive type artifacts can create duplications of various vascular structures, just like they can gray scale structure.

This is the SMA that's duplicated.

The assumption is that the linear elbow here between the rectus muscles acts like a refractive, it causes the refraction of sound and gives you two re turning signals and duplicates.

The SMA, it can happen in the aorta, of course, it can happen in the bladder and elsewhere.

Flash Artifact

Flash artifact is a result of the fact that when you do set up a preset in your doppler, you are automatically choosing a level over which the gray scale will be written.

Where over which color will overwrite the gray scale.

In other words, there's two fundamentally, there's two flow two flow settings that you're gonna cope with.

One where there's a high flow setting, in which case you want to give priority to the gray scale.

'cause like in a carotid, you don't wanna overwrite the plaque.

Then there are low flows states where you want to give priority to the color.

'cause you wanna see any thread of color, so like ovaries or testicles or things like that where you wanna see all those little threads of color.

The priority will be given to the color.

The presets determine where that color right priority is written.

Here it is in this particular manufacturer, you have this line here, and that line tells you the level of the gray scale over which color will overwrite the gray scale.

Here's that.

What that means is that if that's not set correctly, you can have color overriding the gray scale and indicating flow like we have here, when in fact there is no flow because this person is a CF patient with macrocystic change in the pancreas.

There is no flow, but it's a mis adjustment of the color right priority.

Here's another example. Two minutes, okay?

The flash artifact where movement of the heart causes color to be written within the gray scale and can look and make a abscess look like it is in fact a flow containing structure.

Twinkle Artifact

Twinkle is, we all know about that twinkle is quite a popular artifact.

It's seen around areas with highly reflective and roughened surface.

It's caused by a phase shift between a high part of that roughened surface and a low part.

All that all phase shifts are not motion, but it is sufficient that we will see that twinkle artifact indicating the site of a roughened surface.

It's useful in identifying stones in the kidney as well as stones elsewhere, like here in the gallbladder.

Radiation Force and Perivascular Color Artifact

I will just talk a little bit about radiation force in relation to some tips.

Some shunts, you can get movement of the shunt caused by the pushing pulse of the doppler, and you'll get some artifactual flow around the edges of stent.

Here's one. And so this is due to the force of the sound beam itself, which is of course a pressure wave.

All right, and then perivascular color artifact just merely indicates that any motion's gonna be color encoded.

Things where the soft tissue is vibrating will be color encoded and it can identify areas where there's an abnormality, like a fistula.

Here's a fistula here.

This perivascular color artifact is sort of the sonographic thrill.

All right, thank you very much.