Ultrasound image optimization

Fix the image, don't just press buttons

A bad image does not improve when the operator presses buttons at random; it improves when they understand what is wrong and change one cause at a time^[1,4]. Each symptom on screen points to a control: is the target too small? Think depth. Is the image all bright or all dark? Think gain. Is the deep field dim but the superficial field fine? Think TGC. Is the target edge blurred? Check focus, frequency and angle.

The goal of this chapter is straightforward — turn a confusing screen into a documentable image, in the sense discussed in how to think through an ultrasound exam. Doppler is left out: before color, spectrum and measurements, the B-mode must be good, which is what the essential Doppler chapter assumes.

Before touching the buttons, find the structure

Many images get worse because the learner tries to brighten a screen where the structure is not even well positioned. The correct order protects against this: adjust patient and window, choose transducer and preset, find an easy anatomical landmark, center the structure — and only then adjust depth, gain, TGC and focus^[3,5].

The reason is physical. If the window is poor, high gain only adds noise. If the probe does not penetrate, TGC does not create real echo. If the plane is oblique, a pretty image still yields a bad measurement.

Depth

Depth is the length of the path the machine displays. Too much, and the target is small and you waste screen on tissue that does not answer the question; too little, and the target — or the anatomy just below it — is cut off. The rule of thumb is to start wide to orient yourself, then move closer^[1,5].

Before saving, confirm: is the target whole? Is there enough anatomical context? Is there useless background? Is the structure centered? Does the chosen depth let you save a documentable image?

Overall gain

Gain is amplification. On screen it looks like brightness, but the real effect is larger: it changes how you see texture, margin and content — to the point that a simple cyst can look filled if the gain is too high^[2,5].

There is no magic number: the setting depends on the target, the patient, the window and the machine.

TGC — depth compensation

TGC (time gain compensation) corrects the brightness difference by depth, because deep echoes arrive weaker — they traveled farther and suffered more attenuation. A safe way to use it: set overall gain first, look at the near field, the mid field and the deep field, correct small imbalances with the TGC, and stop before creating artificial brightness bands^[2,3].

TGC should not mask shadowing, gas, bone or an inadequate probe. When the deep image is bad due to a penetration limit, amplifying the background makes the screen brighter, but not necessarily truer.

Focus

Focus is the region where the beam is narrowest, and lateral resolution tends to be best there; on many machines it appears as an arrow or marker on the side of the screen. Place it at the target or slightly below it: too high, and the region of interest loses lateral sharpness; too low, and you deliver resolution where it is not needed^[2,5]. Avoid many focal zones when the image needs fluid motion — several focal zones reduce the frame rate, which matters in dynamic exams.

Frequency and penetration

High frequency shows detail; low frequency reaches deeper. This trade-off is one of the first things a learner must accept, and it follows from the transducer physics seen in planes, orientation and movements^[1,2].

The practical question is always the same: do I want to see detail that is close, or do I need to cross more tissue?

Harmonics and refinement modes

Once the basic B-mode is set, some modes can improve the image. Tissue harmonics and compound imaging, for example, reduce noise, improve margins or make interfaces clearer in certain scenarios. Treat them as a test, not a crutch: view the image without the feature, turn the mode on, compare edge, contrast and penetration, and keep it only if it became more useful for the exam question^[3,6]. If the mode makes everything pretty but hides the deep structure, it did not help — in ultrasound, pretty is not enough.

Artifacts that help interpretation

Not every artifact should be erased; some help recognize what the sound met^[1,6]. Acoustic shadowing appears behind structures that block or strongly attenuate the beam — bone, calcification and many stones — and, when it matches the image and context, strengthens interpretation. Posterior acoustic enhancement appears behind structures that let sound pass with little loss, such as fluid content, and, together with shape, wall and absence of internal echoes, helps read simple cysts.

Artifacts that mislead

Artifact also deceives: it can create an image where there is no structure, hide anatomy or make normal tissue look diseased. Before calling something pathology, run the technical test — change the angle, check whether the structure appears in another plane, lower the gain if there is noise, adjust depth and focus, look for edge shadowing, reverberation or mirror artifact, and compare with the contralateral side when it makes sense^[1,6]. If the finding disappears with a small adjustment, it was probably technique; if it persists across coherent planes, it deserves documentation.

Anisotropy

Anisotropy is a classic musculoskeletal pitfall: tendon, ligament, nerve and even a needle look darker when the beam does not strike perpendicularly, mimicking a lesion. The bedside test solves it — keep the structure centered, rock (heel-toe) slowly, fan the beam to make it more perpendicular, watch whether the dark area disappears, and confirm in short and long axis^[1,7]. If it disappears, it was artifact; if it does not, the finding gains relevance and must be documented carefully.

When to change the transducer

There is a point where adjusting the controls stops working, and the answer becomes changing the probe, changing the window or acknowledging a technical limitation^[1,3]. Signs that the time has come: the linear probe does not penetrate to the target; the convex loses detail on a superficial structure; the field of view does not cover the anatomy; the intercostal window needs a smaller footprint; high gain only adds noise; a lower frequency still is not enough; the image stayed technically limited despite correct adjustment. Changing the transducer is not failure — failure is insisting on a bad image and writing as if it were certainty, an error the normal report chapter addresses in detail.

30-second adjustment checklist

1. Is the target on screen?
2. Is the window the best available?
3. Does the depth keep the target at a useful size?
4. Does the gain preserve texture?
5. Does the TGC balance brightness by depth?
6. Is the focus on the target?
7. Does the frequency match the depth?
8. Do harmonics or compound really help?
9. Could the problem be an artifact?
10. Do I need to change the transducer or record a limitation?

Window → depth → gain → TGC → focus → frequency → artifacts → documentation.

If you are lost, go back to the target structure. The best image is not the brightest one: it is the one that best answers the exam, with honest technique. Real cases to train the eye are in the case library.