3D System matrix recovery based on iterative up-and-down sampling super-resolution network in magnetic particle imaging

Magnetic Particle Imaging (MPI) is a promising technique for mapping magnetic nanoparticle distributions within biological tissues. The reconstruction process, which relies on the system matrix (SM), is crucial for accurate MPI imaging. However, the time-consuming nature of SM measurements often requires repetition whenever there are changes in scan parameters, particle types, or environmental conditions. In this study, we proposed a 3D iterative up-and-down sampling super-resolution network(3D-ISSRnet) to solve the SM recovery problem by transforming it into a deep learning super-resolution reconstruction problem in the image domain. This network employs an iterative up-and-down sampling structure to effectively capture deep relationships between low-resolution (LR) and high-resolution (HR) image pairs. Additionally, it incorporates a pyramid pooling module to optimize the utilization of global and local contextual information, while implementing a Dense connection to stream-line model complexity. The experiments conducted on OpenMPI data demonstrate the excellent SM recovery capability of our proposed method. We believe that this research will enhance the practicality of MPI in biomedical applications and contribute to the future advancement of MPI technology.