Experimental Evaluation of an Iterative Reconstruction Method for Time-Correlation Magnetic Particle Imaging

We have previously proposed an image reconstruction method to improve the image resolution in magnetic particle imaging. In that method, the image is reconstructed based on time-correlation information between the observed signal from magnetic nanoparticles and the corresponding system function. However, image blurring and image artifacts were not sufficiently improved under this method. Therefore, we proposed a new iterative image reconstruction method, in which the reconstructed image is iteratively estimated by correcting the difference between the analytically calculated signals from the reconstructed images and the observed signals during iteration. In this paper, to confirm the validity of this new proposed method, we performed experiments using the 3-D prototype system and numerical simulations based on the prototype system setup. The results achieved by the prototype system experiments showed that the image blurring was suppressed and the image quality improved as compared with the previous time-correlation reconstruction method.