Modified Jiles–Atherton Model-Based System Matrix Generation Method for Magnetic Particle Imaging

The system matrix (SM) is an important component of magnetic particle imaging (MPI) because it describes the mapping between the particle concentration and voltage signal. The acquisition of SM predominantly relies on measurement-based calibration. However, the measurement procedure is time-consuming and susceptible to noise interference, leading to a decrease in reconstruction quality. The need for measuring the SM is because existing MPI forward physical process models are not sufficiently accurate for model-based SM generation. To address this problem, we propose a model-based SM generation method based on the modified Jiles–Atherton (MJA) for accurate and fast modeling of the MPI forward physical process (MJA-SMGM). In the proposed method, the MJA accurately describes the dynamic magnetization responses of superparamagnetic iron-oxide (SPIO) nanoparticles. The MJA model considers the relaxation effect to improve accuracy and reduce complexity to accelerate solving. Additionally, time-harmonic electromagnetic field analysis for inhomogeneity and the transfer function of the receive signal chain are included for SM generation. The proposed MJA-SMGM method can achieve fast and accurate SM generation, thereby improving the quality of reconstructed images. Compared to the complex Fokker–Planck method, MJA-SMGM is approximately 70 times faster and reduces error by 30.2% according to the results of homemade MPI scanner experiments. Additionally, the proposed method improves image quality by 2.2 times in terms of the signal-to-noise ratio. These results indicate that MJA-SMGM has the potential to improve the applicability of SM methods in various MPI scanners.