Sensitivity Matrix Reconstruction in Ultrasound Transmission Tomography Using Singular Value Decomposition

Abstract

The main purpose of this paper is to present a method for determining the approximation of the sensitivity matrix in a computed tomography problem. A method based on multiply regression and singular value decomposition was proposed. The study was carried out on 5000 simulated cases for ultrasound tomography. The authors first checked the reconstruction quality obtained by the proposed method. Then these results were compared with reconstructions obtained using the elastic net method. The reconstruction times obtained by the different methods were also examined. The analyses conducted in the paper showed that the proposed method is much faster than the Elastic net method, while the reconstruction quality is comparable.