Strain estimator based on analytical phase tracking

This paper presents a new analytical phase tracking (APT) based strain estimator, suitable for imaging elastic properties of biological tissues under quasi-static external compression. This method utilizes the Hilbert transform to obtain a sequence of 2D analytical RF signals versus compression time and extracts the phase in discrete form to produce a wrapped phase matrix. It subsequently creates zero phase bands using Connected Component Labeling (CCL). Displacement versus range is then estimated based on the difference between the initial and end axial indexes of zero phase bands and converted to strain value. Simulations and experiments with an inhomogeneous phantom demonstrate that APT is capable of reliably estimating tissue displacement and strain over a larger range of deformation than the standard time domain cross correlation (SCC). Statistical simulation assessments at compression level 3% and 6% with SNR 20 dB yield an average strain error of 10% for APT and 15% for SCC. Computation efficiency analysis also shows that APT is roughly 20 times faster than the SCC with comparable or even better accuracy.