19.3 Artifact Recognition: Reverberation, Shadowing, Mirror, Refraction, and Aliasing

Test the scanner's assumptions

An artifact is displayed information that does not accurately represent anatomy or physiology. The scanner assumes sound travels in a straight line at 1,540 m/s, reflects once, returns directly, attenuates uniformly, and originates within the main beam. It then places each echo at depth = speed × round-trip time / 2. Violating one or more assumptions creates a predictable false location, duplication, absence, or Doppler display. Do not simply label a suspicious finding “artifact”; identify its mechanism and show how an alternate acquisition changes it.

Start with a verification sequence: reproduce the finding through several cardiac cycles, change window and beam angle, optimize gain and frequency, and compare orthogonal planes or prior studies. True anatomy preserves coherent relationships as the transducer moves. Artifacts often remain tied to a strong reflector, beam direction, depth, or machine setting.

Recognize axial artifacts

Simple reverberations often form parallel equally spaced lines; complex reverberation can create dense comet-tail or device-related patterns. A reverberation in the LA or aorta can mimic thrombus, membrane, or dissection. It should not maintain a genuine independent attachment in orthogonal views. Shadow is missing information, not proof that no lesion exists behind a prosthesis or calcium. TEE, an alternate transthoracic window, CT, CMR, or appropriately optimized enhancing-agent study may be needed when the hidden region matters.

Mirror images are commonly created near strong interfaces such as diaphragm-lung or pericardial boundaries. Because the system assumes one straight round trip, it places the duplicate too deep. The mirror and real structure share timing and motion.

Recognize lateral displacement and beam-width effects

Refraction bends sound at an oblique interface where propagation speeds differ. The scanner still draws the echo along the original beam line, so one structure may appear laterally shifted or duplicated. Fat-muscle interfaces and the lung edge can act as refractors. Move the transducer, use a more perpendicular angle, and track landmarks; a true second structure persists from independent windows.

Side lobes send weaker energy away from the main beam. A strong off-axis reflector can be assigned to the main beam and displayed inside a chamber. Beam-width and slice-thickness artifacts include structures lying within the finite 2-D or elevational sample even when they are not in the assumed central plane. These mechanisms can create false intracavitary densities or color outside a true jet. Narrow the sector, change focus and plane, reduce excessive gain, and require orthogonal confirmation.

Separate Doppler aliasing from false flow

PW and color Doppler sample at a finite pulse repetition frequency. The Nyquist limit is PRF / 2. When the Doppler shift exceeds that limit, spectral PW wraps to the opposite side of the baseline and color abruptly reverses or forms a mosaic. Aliasing correctly signals that the sampled shift exceeds the setting; it does not by itself specify stenosis severity or prove turbulent flow. High transmit frequency, high velocity, excessive depth, and unfavorable scale increase susceptibility.

To reduce PW or color aliasing, raise PRF or velocity scale, decrease depth to permit higher PRF, use a lower transmitted frequency, or move the sample to a lower-velocity site when localization is the goal. Shift the baseline to display more velocity in one direction, recognizing that this reallocates the display and does not change the physical sampling limit. Use CW Doppler for the true maximum high velocity; CW does not alias but loses range specificity. Never angle-correct or trace a wrapped PW envelope as though its top were the true peak.

Spectral mirror or crosstalk duplicates a strong Doppler signal across the baseline, often from excessive gain or near-90° incidence, and can mimic bidirectional flow. Lower spectral gain, optimize wall filter and alignment, and obtain a different window. Color blooming from excessive gain enlarges jets beyond their true borders; color shadow removes data behind calcium or prosthesis; color splay or side lobe places color outside the main beam. A true jet has an anatomic origin, consistent timing, and reproducible direction.

Mitigate without manufacturing normality

Settings should test a finding, not erase it. Lowering gain until all echoes vanish does not prove artifact, and raising the color scale until regurgitation disappears does not prove normal flow. Save the original and optimized clips when a potential artifact changes interpretation. State the obscured region and the residual uncertainty. Escalate when a suspected thrombus, dissection, prosthetic complication, shunt, or critical velocity cannot be separated from artifact. A defensible study explains the artifact mechanism, demonstrates the acquisition response, and avoids both false-positive anatomy and false-negative reassurance.

Mitigation is a controlled test

Change one relevant variable—window, angle, frequency, gain, depth, PRF, or modality—while preserving anatomy and timing. Erasing a finding with extreme settings is not evidence that it was artifactual.