A 2D phase zero algorithm for estimation of displacement in ultrasound elastography based on the optimization of initial value for iteration

摘要

Ultrasound elastography (UE) is an imaging technique based on tissues’ elasticity. UE based on gradient typically assumes that movement only occurs along the propagation direction of an ultrasound beam. However, when the target tissue is compressed by a transducer, lateral displacement of the scattering body occurs due to uniform hardness within the tissue. Before and after compression, the signal cross-correlations within a specific window along the scan lines may be very low. This is highly likely to result in large errors or even failed calculation of cross-correlated phase angles within this window. We proposed a modified 2D phase zero algorithm for the estimation of displacement. In this method, the initial value with threshold for iteration was optimized in order to decrease the errors result from lateral motion. Then, to verify the effectiveness of this algorithm, a contrast experiment on this new and conventional method was designed. The experiment indicated that this algorithm could effectively prevent failed calculation of elasticity due to lateral displacement of the scattering body. It could stably and efficiently reduce the errors in displacement estimation and enable a more accurate imaging of the target hard mass, like signal to noise of elastography ( S N R e ) and contrast to noise of elastography ( C N R e ).