Image Reconstruction with Relaxation Estimation for Non-Cartesian Magnetic Particle Imaging

Abstract

Magnetic Particle Imaging (MPI) is a relatively new biomedical imaging modality that images the spatial distribution of superparamagnetic iron oxide nanoparticles. In MPI, an AC excitation field is applied to induce nanoparticle signal. The magnetization response of nanoparticles to this excitation field is delayed due to the relaxation effect, which in turn can cause a significant level of resolution loss in the MPI image. In this work, a back-and-forth scanning scheme is proposed for non- Cartesian trajectories in MPI to directly estimate the relaxation time constant from the acquired MPI signal. In addition, using the estimated time constant from both scans, deconvolution of relaxation effects from the MPI signal followed by a gridding reconstruction to obtain a Cartesian MPI image is proposed. The simulation results obtained using realistic parameters show that the resolution of the reconstructed MPI image improves significantly with the proposed method, and that the image quality closely matches that of the ideal image without relaxation effects.