Distributed aberration correction in diagnostic ultrasound by time-delay computation from sound speed estimates using common midpoint gathers

We describe a method of distributed aberration correction from local sound speed estimates of tissue. In this case, compensated beamforming delays are computed using the eikonal equation to account for refracted paths of the ultrasound waves based on an estimation of the local sound speed of the tissue. Signals acquired from a multistatic synthetic aperture sequence are used for image reconstruction in order to refocus the ultrasound beam at every image point without re-transmission. We estimate sound speed using the common midpoint gathers (CMGs) of the synthetic aperture signals, which have high signal coherence that can be used for accurate time measurements. These time measurements are used in overlapping ray paths traced onto a discrete grid to solve an inverse ultrasound computed tomography problem. We separately utilize absolute delay measurements at each depth (to capture axial variations) together with differential time-delays between CMG gathers (to capture lateral variations). For generalized sound speed reconstruction, we balance the contributions of both terms with regularization. In simulations, local sound speed was reconstructed with an rmse of 5.3 m/s in four-layered media and 6.0 m/s in focal lesions. In multilayer propanol/agar phantoms, local sound speed was reconstructed with an rmse of 5.0 m/s.