Physics

Aliasing Phenomenon

Aliasing phenomenon is a concept that causes a lot of uncertainty for many of us. We all have a rough idea of what it is, until we have to explain it to a trainee sonographer doing their best impression of a 3 year old [insert annoying voice here] “… But why???”.

This week’s post was requested by Bec from South Australia. Thanks for jumping straight in with the tricky stuff!

Aliasing is an imaging error which occurs due to under-sampling. The ultrasound system is trying to image an event that is occurring faster than the rate we are sampling it, and as a result the system is uncertain about the direction of the signal and displays this as heading in the opposite direction.  We see examples of aliasing in both colour and spectral Doppler, where the velocity exceeds the Nyquist Limit and the image displayed would suggest flow is heading in the opposite direction.
Cfi_aliasing.fw           alias_pw.fw
Aliasing phenomenon is not unique to ultrasound. I can remember watching the cartoon, SpeedRacer and being fascinated that the wheels of SpeedRacer’s car would start to go in the reverse direction as he sped off.
speed-racer-wheel
This is also known as the wagon-wheel effect, as it was often seen in old western movies. The old movie cameras had a very low frame rate compared to modern movie cameras. This was not an issue for relatively slow moving events, such as people walking around, but the wagon-wheels turned a lot faster than the camera was taking the frames for the movie. At slow speeds, the wheel looked to be turning normally. As the wheel sped up, the image sped up to until it reached the same speed as the frame rate of the camera. At this point the wheel may look to be almost stopped on the film. With increasing speeds, the wheel then appears to go backwards.

Strobe_2

In the first window (continuous imaging) we see the line heading in a clockwise direction. At a slower sampling frequency (1 frame/revolution) the line appears to be stationary. At an even slower sampling frequency, the line appears to be heading anti-clockwise (opposite direction to reality).

This optical illusion became characteristic of speed, hence the artists drawing the effect for SpeedRacer.
We also see similar optical illusions created by aliasing with airplane propeller blades and strobe water fountains. 
Check out these cool examples of aliasing:

What does Speed Racer have to do with the display of our colour Doppler echo images???
The Nyquist limit is determined by half the pulse repetition frequency.
Nyquist Limit = 1/2 PRF
The PRF (and therefore Nyquist limit) become the old-style movie camera in our story. The moving blood cells are the wagon wheel spokes. Remember colour is a pulse wave Doppler technique… a pulse is sent and the system then waits for the signal to return before it sends another pulse. If a second pulse is sent before the the first one is received, the machine doesn’t know which pulse the returning echo has come from…so it waits a designated time for the echos to return based on the depth of the image. If we are pulsing something in the near field, we don’t need to wait as long for the echo to return as we do if we are pulsing something farther away.
The greater the depth, the lower the PRF (needs to wait longer for the signal to return) and therefore the lower the Nyquist limit (maximum velocity that can be displayed unambiguosly).
So if we have a high Nyquist limit, and a low blood flow velocity, the blood flow direction is displayed correctly. As the blood flow speeds up, and exceeds the velocity of the Nyquist limit, the system is not sampling often enough to be certain of which direction flow is occuring in and therefore it is displayed erroneously as “wrapping-around” the scale and appearing as if the direction has been reversed. The ultrasound system has experienced the same optical illusion that we see looking at the wagon-wheels.
We want to display direction of flow correctly… How do we avoid aliasing?
It may be important to display the information on screen without aliasing. For example, aliasing can make visual estimation of regurgitant lesions difficult (often overestimates severity) – try estimating amount of mitral regurgitation with your colour scale set to +0.2/-0.2 m/s….
The following techniques can be used to minimise the amount of aliasing.
  1. Increase the Nyquist limit (increase velocity scale)
  2. Shift the baseline To increase the Nyquist limit in a particular direction
  3. Activate high PRF mode (separate post on HPRF coming soon…)
  4. Change Doppler angle to minimise the Doppler shift. I can’t really see any point to this…we have now substituted one error (range/direction ambiguity) with a different error (underestimation of the Doppler shift therefore underestimate velocity)
aliasing_fix

Increase colour scale or shift colour baseline to minimise aliasing effect.

Shifting the scale and baseline is commonly used to optimise PW Doppler

Shifting the scale and baseline is commonly used to optimise PW Doppler

Can aliasing be useful in ultrasound? Yes. We often use aliasing to help identify regions of flow acceleration or turbulence. This may be important to help try and locate a narrowing or obstruction. We also take advantage of how clearly aliased flow can be seen to identify ventricular septal defects by CFI. In other words the contrasting colour of the aliased flow to the normal flow draws attention to the region, making it easier for the sonographer to appreciate. We all know how difficult seeing an ASD with low velocity flow (no aliasing) can be!
Can we add the velocities to calculate the “correct velocity”? No. If you need to measure a velocity, then take measures to avoid aliasing. There is a portion of the signal “lost” as aliasing occurs. This is analogous to the wagon wheel appearing like it isn’t turning when the revolutions matched the cameras frame rate. We simply don’t know how much signal is lost as it wraps around.
Does understanding the physics detract from the magic of Speedracer? Absolutely not! In both SpeedRacer and during our echo examinations, aliasing phenomenon simply means that something is going fast. This is exciting in the cartoon because you knew that SpeedRacer wasn’t just going for a Sunday drive to the shops, and it is really useful data in our echo exams, alerting us to the higher velocity blood flow present (and possible pathology).
Hope that helps, Bec… at the least you will probably get the SpeedRacer theme stuck in your head for the rest of the week! Enjoy!!

Previous Post Next Post

13 Comments

  • Reply Bec Perry November 12, 2014 at 9:31 pm

    Thanks Tony – what a great explanation, even with the nerdy references to speed racer! Looking forward to the next blog post.

    • Reply echo.guru November 12, 2014 at 10:16 pm

      Blaspheme!!! SpeedRacer is never “nerdy”!!!!! Will give physics a rest for a while…next post will jump back to the basics and look at “Tips for obtaining on-axis images”. Stay tuned!!

  • Reply Luke November 16, 2014 at 10:29 pm

    Thanks Tony,
    Physics is Phun!

  • Reply Lisa November 24, 2014 at 5:20 am

    Now we can refer our trainees here!

  • Reply Dr.fuad kurdi September 29, 2016 at 7:06 pm

    Thanks alot

  • Reply Luke Platt April 14, 2017 at 8:49 am

    Hey Tony, just read your explanation of Aliasing and the Nyquist Limit……..best explanation i have heard….thanks very much and go Speed Racer!

    • Reply echoguru April 14, 2017 at 10:03 am

      Thanks Luke. All my students now are too young to even know speed racer, so it might be time for a revamp! Cheers

      • Reply Tara Curry June 3, 2017 at 12:21 am

        I LOVED Speed Racer as a kid….had a crush on him, so I would say keep him on!
        But then, I agree, the students may not know him.
        Great explanation though, easily understood, good for Physics review!

        Thanks!

        • Reply echoguru June 3, 2017 at 9:14 am

          Thanks Tara. I didn’t have such strong feelings but he was pretty awesome! Speed racer can stay… 🙂

  • Reply Kristen September 1, 2017 at 8:59 pm

    Hi Tony! Can you explain why decreasing the color scale to look for low flow (we decrease when putting color on the IAS from the subcostal view) is helpful? Why does it help to bring out lower velocities with color? Also, do you recommend maxing out the color scale for high velocity flow (ie: MR, AI, VSD, etc)? What do you typically recommend for optimizing color scale?

    • Reply echoguru November 1, 2017 at 3:35 am

      Hi Kristen
      Decreasing the color velocity scale (Nyquist limit) allows for the range of available colors to be spread out across a much lower range. This allows low velocity flow to be more easily displayed. At a high Nyquist limit, low velocity flow would be displayed as a dark blue (or red) hue…or may be completely absent as it is filtered out. Reducing the velocity scale would mean that we can better appreciate the low velocity flow, but the point at which aliasing occurs is lower too. Let’s look at the left atrium as an example, in a patient with mitral regurgitation (MR). If we put color Doppler at a maximum velocity scale over the LA, we might see the MR clearly, but not the pulmonary venous flow. Reducing the scale will increase the colour display of the pulm veins, but it will also mean that the aliasing velocity is reduced and the MR will appear a lot worse that it really is.
      The current guidelines are to set the CFI scale between 50-70cm/s for the assessment of valvular regurgitation. This is somewhat different to earlier guidelines suggesting that we should be at max scale for qualitative assessments. This potentially has big implications for serial assessment especially if the first study was at a high velocity (~1m/s) and the second study has been imaged at 0.5m/s. My strong suggestion is that if you are going to start implementing a less than max approach, and the patient has had a previous study, then you should prob store a second clip at a similar velocity to the first time to avoid falsely over-calling a change in severity. Obviously this is not the only parameter we look at when assessing change, but you would be surprised how often everything else is thrown away just because the colour looks different to last time.
      I image my high velocity lesions at maximum…but in the vast majority of patients the Nyquist limit would be between 0.6 and 0.8cm/s anyway due to the depth of the signal from the probe.
      Hope that helps.

      Tony
      PS sorry for the really long delay in reply…the website had some downtime and a few things dropped off in the re-build. I lost track of your comment altogether.

  • Reply The "Sound" of Performance Monitoring, Part 2: Aliasing July 17, 2018 at 12:31 pm

    […] deal with “Purple Haze,” Ultrasound can have flows look like they’re going in the reverse direction, MRI has to battle “wrap-around,” and Mass Spectrometry has similar […]

  • Reply Stephen August 29, 2018 at 3:59 pm

    Thank you very much for this straightforward explanation. It’s just what I needed.
    Go Speed Racer!

  • Leave a Reply

    %d bloggers like this: