Imaging performance of thin-film disk particles tailored for optimal MPI signal generation

Abstract

The imaging performance of a Magnetic Particle Imaging (MPI) scanner is fundamentally constrained by the magnetic characteristics of the employed tracer. The achievable spatial resolution depends on the magnetic particle used, particularly on its magnetic moment's reaction to the applied oscillatory field, which, for commonly used superparamagnetic particles, is limited by their Langevin-type magnetization behavior. An optimal nanoparticle would demonstrate a high magnetic moment and a step-like magnetization response to the magnetic field oscillations. This sharp response leads to a narrow point spread function and enhances the amplitude of the higher harmonics. Thus, higher harmonics, which encode the fine spatial resolution, become detectable above measurement noise.

In this study, we utilized magnetic disk particles fabricated via a top-down approach. With this method of fabrication, the optimization of the particles’ magnetic properties is not constrained by chemical synthesis limitations. We then evaluated the imaging performance of these particles by employing multi-dimensional (one-, two-, and three-dimensional) excitations and determining the system matrices using a multi-dimensional magnetic particle spectrometer. We then benchmarked them against Perimag, the established gold-standard MPI tracer. The system matrices of the disk particles revealed significantly stronger signals, especially noticeable in the high-order harmonics. Furthermore, the structure of these one- and two-dimensional matrices closely resembled Chebyshev polynomials, without the typical vignette effects visible by wet synthesis particle systems. Finally, an estimation for the achievable resolution improvement is given by using the sensitivity and noise level of a preclinical MPI scanner within a simulated software phantom.