Speaker
Description
One of the main challenges in algebraic reconstruction techniques (ART) for computed tomography (CT) is the generation of the matrix of the weighting coefficients, also called system matrix (SM). It is well established that in theory the exact SM entries are computed as the relative sub-voxel volumes covered by X-rays towards a detector element. However, due to high computational efforts, in practice approximations based on line integrals are frequently used. In this work, we propose a novel method for the computation of the exact system matrix for two- and three-dimensional cone-beam flat-detector CT. The method relies on the decomposition of the cone-voxel intersection volumes into subvolumes that contribute to distinct detector elements and whose contributions to the system matrix admit analytical expressions that can be evaluated without costly iterative subroutines. We demonstrate that the reconstructions obtained with the proposed method are superior to those obtained with common line-based integration approaches with numerical experiments on synthetic CT data.