Signal Processing: Demodulation, Rejection & Pre/Post-Processing

Overview: Pre-Processing vs. Post-Processing

Before a returning echo signal becomes a pixel on the screen, it passes through a series of signal-processing steps. SPI groups these into two broad categories:

Nearly every function in this section can be classified as pre- or post-processing, and SPI regularly tests which category a given function belongs to.

Demodulation and Rejection

Demodulation converts the raw, high-frequency radiofrequency (RF) echo signal into a usable video signal. It has several sub-steps, including amplification (Section 7.2), compression (Section 7.3), and rectification (converting the RF waveform, which has both positive and negative voltage swings, into an all-positive envelope that represents echo amplitude over time). Demodulation happens early, before the signal reaches the scan converter, and is therefore a pre-processing function.

Rejection (sometimes called suppression) is a receiver function that eliminates low-amplitude signals below a chosen threshold — typically electronic noise or very weak scatter — so they do not appear on the display as unwanted low-level gray "haze." Raising rejection cleans up the image by removing faint noise, but rejection set too high can also discard genuine weak echoes; it is generally classified as pre-processing since it acts on the signal before scan conversion.

Read Zoom vs. Write Zoom

Zoom magnifies a region of interest, but SPI distinguishes two very different mechanisms:

• Write zoom (also called "true" or "research" zoom) is applied before the scan converter writes the image into memory: the sonographer selects a region of interest and the system re-scans that region using a higher scan-line density, actually acquiring more data points to fill the zoomed region. Because it changes what data is written to memory, write zoom is a pre-processing function, and it genuinely improves detail/resolution within the zoomed region.
• Read zoom (also called "pan" zoom) is applied after the image is already stored in scan-converter memory: the system simply magnifies the existing stored pixels, enlarging them on the display. Because it only reads and re-displays already-stored data, read zoom is a post-processing function, and it does not add new information or improve resolution — it only makes existing pixels bigger (similar to zooming in on a digital photo, which can produce a blocky/pixelated look at high magnification).

Gray-Scale Maps

A gray-scale map (also called a gray-scale curve or transfer curve) is the function that assigns a specific shade of gray to each possible echo-amplitude value after dynamic range compression (Section 7.3). Different preset gray-scale maps emphasize different parts of the amplitude range — for example, one map might favor contrast in low-amplitude (weak echo) regions, while another favors contrast in high-amplitude (strong echo) regions. Because gray-scale maps are typically adjustable on a stored/frozen image, they are classified as a post-processing function.

Persistence (Frame Averaging)

Persistence, also called frame averaging or temporal compounding, is a post-processing technique that blends several consecutive image frames together before displaying the composite. Because random noise varies frame-to-frame while real anatomic structures remain consistent across frames, averaging frames together smooths out noise and speckle, producing a less "grainy" image. The trade-off is temporal resolution: high persistence blurs rapid motion (because it blends frames spanning a longer time interval), so persistence should be reduced when imaging fast-moving structures such as a beating fetal heart or valve leaflets.

Edge Enhancement

Edge enhancement is a processing technique that increases the amplitude difference at boundaries between adjacent structures with different echogenicity, making borders/edges appear sharper and more distinct. It works by exaggerating amplitude transitions rather than adding new anatomic information, improving the subjective visibility of a boundary such as an organ capsule or a mass margin.

Frequency Compounding

Frequency compounding transmits and/or processes multiple different frequency bands (rather than a single frequency) from the same location and combines the resulting images into a single composite frame. Because each frequency band produces slightly different speckle and interference patterns, averaging across bands reduces speckle noise and improves overall image uniformity — a mechanism distinct from spatial compounding (Chapter 8), which instead combines images acquired from multiple beam-steering angles.

Summary Table: Classifying Each Function

Mastering the pre-/post-processing split, and especially the read-zoom-vs-write-zoom distinction, directly covers Domain 3 tasks 3.E (magnification), 3.M (gray scale, frequency compounding, persistence), and 3.N (edge enhancement).