Background signal suppression using a transformer-based masked autoencoder for magnetic particle imaging

Magnetic Particle Imaging (MPI) is an emerging imaging technique that utilizes the nonlinear response of super-paramagnetic iron oxide nanoparticles to generate an image of their spatial distribution. To achieve high-quality MPI images, it is crucial to suppress background noise. In this work, we propose a transformer-based masked autoencoder for learning the relationships between harmonic components to improve noise suppression. Experimental results demonstrate that the proposed method effectively reduces background noise across varying levels.