Adaption of direct Chebyshev reconstruction to an anisotropic particle model

Abstract

Model-based image reconstruction in magnetic particle imaging (MPI) is an alternative to common reconstruction methods relying on a measured system matrix. It avoids the time-consuming measurement process but has the problem of inferior image quality due to the complex imaging chain that has to be modeled. Recently, a direct reconstruction method using weighted Chebyshev polynomials for multi-dimensional MPI has been proposed that operates in the frequency domain and even does not need a simulated system matrix. However, as the underlying model neglects several physical processes, including the anisotropy of nanoparticles, artifacts in the reconstructed particle distribution can occur.

In this work, an adaption of the direct reconstruction method to an anisotropic particle model is proposed. It is shown that the adaption reduces deformations of the reconstructed particle distribution and thus provides one further step towards fast and high quality model-based image reconstruction.