The most fundamental type of frame averaging is “persistence.” This is an adjustable setting present on almost all ultrasound machines, with the exception of some of the current palm-held machines and wireless ultrasound probes, designed as screening tools rather than as full diagnostic ultrasound machines (to view part of my dissertation on handheld ultrasound as a screening tool, visit: pocket echocardiography).
Increasing persistence tends to result in a smoother image, but one which will appear delayed or ‘laggy’ with rapid movement of the target such as a kicking foetus or a heartbeat, or even the panting of a dog or the movement of the operator’s hand. Decreasing persistence results in a much more responsive image, but one which will also appear more ‘grainy.’
Frame averaging can create a more aesthetically pleasing image and reduce the appearance of speckle. The logic behind this being that whether speckle is prominent or not is a matter of wavefront timing rather than the inherent properties of one area of tissue versus the next. Speckle reduction can also make the image more diagnostic by improving contrast resolution. As a post-processing technology, however, it’s not doing anything to discern true reflectors from artefact.
A more advanced method of speckle reduction is spatial compounding. This involves sending separate ultrasound beams from different angles, and then averaging these, rather than simply averaging the same repeated beam from the same angle. This way, if a structure is truly artefactual, it won’t be visible from multiple different angles. Summing of these different returning beams by the ultrasound machine results in an ‘averaging out’ of speckle and an overall reduction in artefact. Like persistence, however, there is a trade-off in terms of frame rate and responsiveness.