Algebraic determination of back-projection operators for optoacoustic tomography

source: © 2018 Optical Society of America

The simplicity and computational efficiency of back-projection formulae have made them a popular choice in optoacoustic tomography. Nonetheless, exact back-projection formulae exist for only a small set of tomographic problems. This limitation is overcome by algebraic algorithms, but at the cost of higher numerical complexity. In this paper, we present a generic algebraic framework for calculating back-projection operators in optoacoustic tomography. We demonstrate our approach in a two-dimensional optoacoustic-tomography example and show that once the algebraic back-projection operator has been found, it achieves a comparable run time to that of the conventional back-projection algorithm, but with the superior image quality of algebraic methods.[Read More…]

Fig. 1 (a) The grid of the image and detector locations used for calculating the model matrix?. The image is divided into ??×?? square pixels with a pixel area of ???? and the acoustic signals are sampled at ?? positions over a line with a distance of?? between them. (b) The image grid on which the projection operator ? is calculated. Here, only a single back-projection is calculated, and the number of pixels in the x directions is increased to ??=??+??−1.

Amir Rosenthal, “Algebraic determination of back-projection operators for optoacoustic tomography,” Biomed. Opt. Express 9, 5173-5193 (2018)