Speed-of-Sound Dispersion Estimation from Pulse-Echo Data

Variations of shear wave speed with the frequency show value as a viscoelasticity biomarker. On the other hand, speed-of-sound dispersion (SoS-D) in soft tissues is weak and has been mainly reported in in vitro scenarios. In this work, we propose a method for estimation of SoS-D based on pulse-echo probes. To minimize wavefront distortion, SoS-D is directly estimated from full-synthetic aperture data. A previously reported spatio-temporal filter allows extracting and aligning coherent waveforms along Common-Midpoint (CMP) gather ray paths. Then a phase spectrogram is calculated with a Short Fourier Transform (SFT) to compute phase delays and average phase velocities in function of the frequency. By combining ray paths from multiple CMP subapertures, an inverse Ultrasound Computed Tomography problem for SoS-D estimation is solved. SoS measurements calculated in the SFT domain are consistent with estimations in the time domain. Phantom results show higher values for biological ex-vivo tissues (>1 m/s/MHz) than urethane phantoms (<0.2 m/s/MHz). In murine NAFLD models, a strong correlation (r = 0.707) is observed between SoS-D and and liver lipid content. The results show feasibility of SoS-D reconstruction and highlight its potential for quantitative tissue diagnostics.