Exact Inversion from Space-filling Trajectories in Cone-beam Transmission Tomography

Murdock Grewar,Glenn Myers,Andrew Kingston
2024-12-06
Abstract:This article introduces a new theory of exact inversion in cone-beam transmission tomography where the source point locus is a 2D surface, 3D volume, or something more complex. We specialise the theory to the case of the cylinder-shaped source locus, and we describe in detail a functioning practical implementation of the inversion algorithm for this geometry. This locus is accessible to CT scanners which include translational and rotational manipulators (e.g. typical helical scanners). This serves as a concrete instantiation of the theory and as a reference implementation which is immediately applicable to many contemporary scanning apparatuses. We illustrate some error characteristics of the algorithm through reconstructions of a simulated dataset. We finish with a reconstruction performed on a set of experimental data acquired with a low-pitch sparsely sampled helical trajectory.
Medical Physics
What problem does this paper attempt to address?