The central slice theorem generalization for a fan-beam tomography

The problems of few-view tomography require sophisticated iterative algorithms which employ a priori information on an unknown object. One of the well-developed algorithms for parallel tomography is the Gerchberg-Papoulis algorithm, which alternately iterates images in Fourier space and in image space. The application of this algorithm in the case of fan-beam tomography is blocked by the lack of the corresponding central slice theorem that connects 1D Fourier coefficients of projections with the Fourier coefficients of a 2D image. In this paper, we formulate the central slice theorem for the case of fan-beam tomography. The use of this modified theorem is illustrated by several numerical examples.