Implementation of ultrasound speckle filters for clinical trial

Abstract

Two methods are described and evaluated for compensating for the signal processing characteristics of ultrasound scanners in an adaptive speckle filter. Precorrection, in which the B-scan is transformed by the inverse of the scanner transfer function, is effective if the algorithm is calculated in full floating point arithmetic (FFP), but integer implementation results in artifacts. Postcorrection, where the speckle reference is adjusted according to its observed variation with the mean amplitude, is equally effective whether implemented in FFP or 16/8-b integer arithmetic.<>