Design and Optimization of a Selection Field Generator for a Human-Sized Magnetic Particle Imaging Head Scanner

Abstract

Magnetic particle imaging (MPI) is currently on the verge of moving from a pre-clinical to a clinical imaging method. In 2019, the first version of a head-scanner and its selection field generator was shown by Graeser et al.. An improvement of the two-coil selection field generator was introduced by Foerger et al., containing 18 iron core coils that can be driven separately. The resulting 9 coils per side are arranged in a 3 x 3 square shape. In this work, this concept is adapted to a human sized scanner examining a set of parameters, including the coil and core size, their material, the current distribution and a division into multiple coils. The setup is simulated in COMSOL 6.1 and optimized with the aim of achieving the target gradient of 0.5 T/m, while minimizing the average power consumption. However, while analyzing the parameters the coil arrangement showed a very promising result in terms of a significant reduction in dissipated power. A MATLAB optimization algorithm, that solves the inverse problem for an optimal current combination for a given target field, reflects the estimated power for each FFP position in the field of view. From this solution a power map can be generated. With this the original sqare- and a new circular arrangement are compared and a reduction of almost half the power consumption can be seen. However, this circular design needs further optimization regarding constructive challenges, but it shows how power-saving selection field generators could look like in the future.