Aliasing in a High-Velocity ICA
Carotid duplex is performed on a 68-year-old with a bruit. Continuous-wave Doppler of the left ICA gives a peak systolic velocity of 1.9 m/s. Pulsed Doppler is then acquired with a 5 MHz linear array at a PRF of 4.5 kHz, 60-degree angle correction, and a 4 cm sample volume depth. The spectral waveform folds: the systolic peak reappears below the baseline with the envelope reversed, and color aliasing appears in the same segment. The baseline is centered and the scale reads plus or minus about 0.7 m/s. The sonographer wants to know whether the 1.9 m/s value can be documented with pulsed Doppler without moving the sample volume deeper.
1. Calculate the Nyquist velocity for the current setting and determine what PRF would be needed to display 1.9 m/s without aliasing at the same angle and frequency.
Read the worked answer
Aliasing threshold: v_max = (PRF x c) / (4 x f0 x cos theta).
v_max = (4500 x 1540) / (4 x 5,000,000 x 0.5) = 6,930,000 / 10,000,000 = 0.69 m/s
This matches the displayed scale of about 0.7 m/s. The true peak of 1.9 m/s is roughly 2.7 times the Nyquist limit, so the peak folds and a value read from the wrapped envelope would grossly underestimate peak systolic velocity. An aliased peak must never be reported as a low velocity.
Required PRF for 1.9 m/s: PRF = (4 x f0 x cos theta x v) / c = (4 x 5,000,000 x 0.5 x 1.9) / 1540 = 19,000,000 / 1540 = about 12,300 Hz
2. Is 12.3 kHz PRF achievable at a 4 cm sample volume depth, and which alternative adjustments would or would not work? Justify each.
Read the worked answer
Maximum PRF set by range ambiguity: PRF_max = c / (2 x depth) = 1540 / (2 x 0.04 m) = 19,250 Hz. So 12.3 kHz is feasible at 4 cm in principle, though many systems will cap duplex PRF well below that. The practical fix is to raise the scale/PRF as far as the depth allows and re-check.
What works:
- Raise PRF (raise scale): directly raises v_max in proportion. Primary answer.
- Reduce transmit frequency to 3 MHz: required PRF falls to (4 x 3,000,000 x 0.5 x 1.9)/1540 = about 7,400 Hz, which is easily reached at 4 cm. Costs axial resolution.
- Continuous-wave Doppler: no PRF, no aliasing, and it is the right tool for a focal high-velocity jet.
What does not work as commonly assumed:
- Reducing the Doppler angle from 60 to 45 degrees makes things worse, because v_max is proportional to 1/cos theta. At 45 degrees, v_max at 4.5 kHz falls to about 0.49 m/s.
- Baseline shift only repositions the display window; it changes what you can see, not the Nyquist limit, and it cannot rescue a peak that is 2.7 times v_max.
- Increasing receiver gain or lowering wall filter changes nothing about the sampling rate.
Increasing the angle toward 70 degrees would raise v_max to about 1.0 m/s, still short of 1.9 m/s, and would degrade velocity accuracy; do not use it as the fix.
3. The peak wraps again even after the scale is maximized. What is the defensible way to document peak systolic velocity, and what should the report say?
Read the worked answer
Document the peak with continuous-wave Doppler (which cannot alias) or with the 3 MHz probe at maximum PRF, and cross-check the two. Note the angle used and keep it at or below 60 degrees so the cosine correction stays defensible.
In the report:
- State the measured PSV and the method and angle used.
- Note that the pulsed Doppler examination was aliased at the maximum achievable PRF and that CW was used for the peak.
- Do not report a value taken from a folded envelope, and do not silently average an aliased number into a ratio.
Judgement: the clinically important number here is the peak velocity used for stenosis grading. A wrapped peak of about 0.5 m/s would push the case from a probable significant stenosis to an apparently normal velocity, which is a far more dangerous error than a slightly noisy CW trace.