- •
Common artifacts related to machine settings are focusing, random noise, aliasing, motion, blooming, and mirroring.
- •
A common artifact related to the surrounding tissue is reverberation.
- •
Common artifacts related to the examiner or patient are motion and pressure.
- •
Spectral Doppler cannot help with mirror artifacts and reverberation artifacts, because they are generated by true flow.
In applying ultrasound to the clinical evaluation of patients with inflammatory diseases, the focus has been on the hyperemic part of the inflammatory process. Gray-scale ultrasound provides valuable qualitative and quantitative information about accessible bone surfaces, tendons, entheses, and joints, and it can visualize bone erosions, synovial hypertrophy, and fluid. Doppler techniques, however, can provide information on the degree of blood flow through a joint, which is considered a better correlate with ongoing disease activity. The need to target Doppler activity was demonstrated in a study of patients with rheumatoid arthritis in clinical remission. Although gray-scale and Doppler findings were both common in patients in apparent clinical remission, Doppler-confirmed activity within a joint was significantly more likely to correlate with progressive development of erosions.
Working with Doppler ultrasound does have pitfalls, and the most important are artifacts. They are often present during the examination and may be related to the machine settings, the interaction from the surrounding tissue, or the examination setup. Doppler artifacts may be sources of misinterpretations, and because they cannot be eliminated totally from the Doppler examination, it is important to understand the types of artifacts that occur most frequently. Preventing and correcting for artifacts can ensure a more uniform interpretation of examinations.
Artifacts Related to Machine Settings
Some artifacts may be accepted when the ultrasound machine is set for slow flow; others may be eliminated by adjusting the machine settings correctly.
Random Noise
Random noise is a common artifact. Random noise is produced in all electrical circuits. When the gain is set too high, this noise becomes detectable in the color and power Doppler circuitry. It is seen as color foci appearing randomly in the ultrasound image. It is easily identified as an artifact because the colors never reappear in the same location as true flow does ( Fig. 2-1 ). Random noise makes it difficult to interpret the image correctly, and to avoid random noise, the gain must be adjusted correctly: The gain is increased until random noise is seen in the image, and it is gradually lowered until only a very few noise pixels are seen in the image. However, if the gain is lowered too much, flow information is lost. Gain is like the volume button on the stereo; if it is set too high, it distorts the sound, and if it is too low, the music is not heard properly.
Aliasing
When used in a rheumatologic context, aliasing for Doppler ultrasound is seldom of relevance, but it is described in this chapter because it is always falsely mentioned as a drawback to color Doppler compared with power Doppler.
Aliasing is a well-known artifact in color and spectral Doppler examinations. The pulse repetition frequency (PRF) is the sampling frequency. The PRF indicates how often the machine emits a pulse and listens for the return signals from the moving blood. If the sampling frequency is too low, aliasing occurs and the incorrect velocity of the blood is displayed. In other words, aliasing arises when the Doppler shift of the moving blood is higher than one half of the PRF, also called the Nyquist limit. When this happens, the signal “folds over,” showing false reverse flow on the display, which is easily demonstrated in spectral Doppler. In Figure 2-2 ( top image ), the spectral Doppler curve folds over showing the top point of the flow curve in the opposite side. In color Doppler images, aliasing is seen as a folding over of colors, thereby showing pixels with an opposite direction from that of the surrounding flow (e.g., red instead of blue and vice versa). In true reverse flow, there is a detectable black line between the opposite directions of flow, because between the areas with opposite directions, there must be an area with no velocity, and no Doppler shift means no signal ( Fig. 2-3 ).
To surmount the problem of aliasing (i.e., displaying wrong colors and thereby wrong velocities), the PRF may be increased to display the correct velocities. But by increasing the PFR, the sensitivity to slow flow is reduced. Another option when working with large vessels where the flow may change due to stenosis is to alter the insonation angle. This is relevant in patients with giant cell arteritis. For spectral Doppler, aliasing may be corrected by moving the baseline up or down (see Fig. 2-2 ).
Aliasing artifact is important only when velocity and direction are important to the examination; it has not proved important in rheumatologic ultrasound. In rheumatology, the presence or absence of colors are relevant, not the direction or the velocity of the detected flow. It is therefore important to keep the PRF low and not to increase the PRF to avoid aliasing. By increasing PRF sensitivity the slow flow is lost (because PRF and wall filter are linked controls and the wall filter will remove low-velocity flow) ( Fig. 2-4 ). The exception to this rule is the application of color and spectral Doppler in vasculitis. Aliasing usually does not occur in power Doppler.
Motion
Motion artifacts are also known as flash or clutter artifacts. The Doppler circuitry detects motion between the transducer and the tissue. When the transducer and the patient are immobile, the only thing moving is the blood, and for practical purposes, the moving erythrocytes that reflect the ultrasound, thereby generating the colors in the Doppler image. Movement of the patient or transducer and movement of the tissue or vessel wall caused by arterial pulsation during imaging produce motion relative to the transducer and produce a Doppler shift ( Fig. 2-5 ). The movements are slower than the flowing blood and therefore produce lower-frequency Doppler shifts that appear as random, short flashes of large, confluent areas of color. Motion artifacts may easily be separated from true flow signals because they are not pulsating and seldom recur in the same place.
One way to avoid low-frequency flash artifacts related to movement of the tissue or vessel wall is to use wall filters. They are high-pass filters, and a Doppler shift must be above a certain threshold to be displayed. However, these filters also remove information from slow-moving blood. Keeping the transducer and patient immobile during examination can eliminate some motion artifacts. When the Doppler is at its highest sensitivity just at or below the noise level, intermittent motion artifacts must be accepted.
Blooming
In a blooming artifact, the color reaches beyond the vessel wall, making the vessels look larger than they are ( Fig. 2-6 ). It is gain dependent, and lowering the Doppler gain minimizes the blooming artifact. However, by lowering the gain, the weakest Doppler signals are lost, and important flow information may be lost. The Doppler gain should be set by looking at random noise, not by looking at blooming artifacts. It is therefore an artifact that must be accepted and that will generate a systematic error in image evaluation. The excess color in the image is not really false because it is generated by flow.
