Signal-to-noise ratio optimized image reconstruction technique for magnetic particle imaging

Abstract

Magnetic particle imaging (MPI) is a new biomedical imaging modality that images the spatial distribution of superpamagnetic iron oxide nanoparticles. In MPI, the amplitude of the excitation magnetic field that causes the time-varying magnetization response of the nanoparticles is restricted by the nerve stimulation safety limits. Hence, the region to be imaged is divided into small sections and scanned as overlapping partial fields-of-view. The nanoparticle signal at the excitation frequency is lost during the filtering process of the direct feedthrough signal induced on the receive coil due to the excitation field. To recover this loss, the overlapping partial fields-of-view are merged via utilizing the continuity and positivity of the desired image. In this work, an image reconstruction technique that merges the partial fields-of-view while optimizing the signal-to-noise ratio is proposed. Accordingly, each partial field-of-view must be weighted by the square of the position-dependent scanning speed. Via extensive simulations at various overlap percentages and signal-to-noise ratios, this work demonstrates that the proposed method overcomes the vertical line artifacts caused by the standard MPI reconstruction techniques and improves image quality.