Multi-variable reconstruction in ultrasound computed tomography with compressed sensing

Abstract

The ultrasound in the brightness mode is a widely used ultrasound imaging technique, but quantitative values in the human body can not be obtained, and the spatial resolution is limited by the beam width. In order to overcome this problem, ultrasound computed tomography (UCT) is used. UCT is a technique to invert the quantitative values inside the human body by receiving scattered waves of ultrasonic waves at various spatially different positions. Using this, it is possible to obtain a quantitative value with good spatial resolution as compared with the brightness mode. However, there is a problem of instability of the solution and large amount of computation for burden. Compressed sensing is used to solve these problems at the same time. In this paper, we use compression sensing to improve the computational complexity and recovery accuracy, and it is beneficial to implement hardware. In this study, three quantitative parameters such as compressibility, attenuation coefficient, and density are obtained by combining UCT and compressed sensing. The recovery time is faster and recovery accuracy increases by obtaining the solution through the greedy algorithm without the solution through the optimization method.