Time-Domain Reconstruction of the Speed of Sound in Ring-Array Ultrasound Computed Tomography with Randomized Super-Shots

TL;DR

This paper introduces a stochastic time-domain reconstruction method for ring-array ultrasound computed tomography, demonstrating comparable image quality to traditional frequency-domain methods but with increased computational time.

Contribution

The paper presents a novel stochastic time-domain inversion algorithm using super-shots and ensembles for ultrasound tomography, expanding the methodological toolkit.

Findings

Time-domain stochastic reconstruction achieves similar image quality to frequency-domain methods.

The proposed method is significantly slower than deterministic frequency-domain reconstruction.

Experimental results validate the effectiveness of the stochastic approach on real data.

Abstract

Ring-array ultrasound computed tomography has recently achieved sufficient maturity for clinical applications like breast imaging. Image reconstruction is achieved with state of art iterative algorithms (full waveform inversion in the frequency domain). In this Letter, we consider a stochastic reconstruction in the time-domain. We introduce the notion of multiple super-shots and stochastic ensembles and test our inversion algorithm on publicly available experimental data. Our results show that image quality of a time-domain stochastic reconstruction may be comparable to image quality of a deterministic reconstruction in the frequency-domain, although the time-domain reconstruction is significantly slower.