A prerequisite for system matrix-based image reconstruction in Magnetic Particle Imaging (MPI) is the acquisition of a system matrix that describes the mapping between the MPI tracer and the measured signal. A common method for its acquisition is a time-consuming calibration procedure, during which the scanner is blocked for other uses. For this reason, a model-based approach that allows the system matrix to be obtained by simulation has great appeal. However, an accurate model that describes the magnetization behavior of the tracer, allows the identification of its parameters, and is computationally feasible has not yet been found.

In an ongoing collaboration with Hannes Albers and Tobias Kluth from the University of Bremen we investigate and refine a magnetization model based on the Néel rotation for the magnetic moments of the particles (see Kluth et al., 2019 and Albers et al., 2022). On the one hand, the identification of the parameters of the model is in focus, since these are unknown a priori, on the other hand, measured 2D MPI system matrices are describe with much higher accuracy than the current MPI models. Moreover, we are also interested in the limitations current MPI models in the context of fluid dynamics (see Möddel et al., 2023).