One shot, one SoS: A real-time, single-shot global speed of sound estimator

Speed of sound (SoS), or the propagation speed of acoustic waves through a medium, is an intrinsic property of human tissue and has emerged as a new biomarker in health diagnostics. Alas, no existing technique has practically demonstrated that the tissue SoS can be robustly measured from a single pulse-echo transmission with an imaging transducer, so incorporating SoS estimation into the ultrasound imaging pipeline remains technically challenging. In this paper, we propose a novel global SoS estimation algorithm that requires only a single steered plane wave transmission for operation. Our single-shot framework derives the SoS estimate by 1) calculating each pixel's pre-beamformed sum of normalized autocorrelation coefficients (SNAC) derived from the time-delayed channel data ensemble for an assumed SoS; 2) constructing a loss metric that is defined as, for different SoS candidates, the negated total SNAC over different pixels; 3) finding the SoS with the minimum loss value. Our single-shot SoS estimator was implemented in real-time (50 ms processing time) on a portable ultrasound research scanner. It was tested in vitro using agar staircase phantoms (SoS range: 1508-1682 m/s) and in vivo using svelte human calves (SoS range: 1573-1589 m/s). All SoS estimates were validated with reference through-transmission measurements. Results show that our framework yielded accurate SoS estimates with a small mean signed difference (MSD) in vitro (0.4 ± 6.5 m/s) and in vivo (3.8 ± 15.7 m/s). When the framework was extended to a 10-angle multi-transmission sequence, its SoS estimation performance was further improved with a smaller MSD (0.2 ± 2.0 m/s). The advent of the proposed single-shot SoS estimator can help advance the emerging use of SoS in tissue characterization and improve other imaging processes that are influenced by SoS, such as beamforming and Doppler estimation.