Two-dimensional magnetic imaging system for evaluating iterative reconstruction method based on time-correlation information

• Main Authors: Tsuchiya, Hiroki, Shimizu, Shota, Hatsuda, Tomoki, Takagi, Tomoyuki, Noguchi, Tomoaki, Ishihara, Yasutoshi
• Format: Conference Proceeding
• Language: English
• Subjects: Image reconstruction; Iterative methods; Magnetic fields; Magnetic particles; Reconstruction algorithms; Solenoids
• DOI: 10.1109/IWMPI.2015.7107053

Magnetic particle imaging (MPI) has been proposed as a new imaging modality and a preclinical MPI scanner was released in 2014. Using this modality, image reconstruction methods that enable the reconstruction of an accurate image under limited hardware resources, e.g., when a relatively lower gradient field strength is applied is attempted. Although the method is an image reconstruction technique based on the iterative Kaczmarz method, it is expected that the noise tolerance in this approach will be higher than that of a regression analysis, since a correlation analysis is used. To confirm the effectiveness of these methods, numerical simulations and experiments using a prototype system were performed. Results show that the use of the iterative methods provides an accurately reconstructed image. However, a clearer image has been obtained through numerical analysis under almost the same conditions. This is believed to be because the system function used for the reconstruction of the experimental data differed from the actual system function. In order to further improve image quality, the authors will attempt to reconstruct an image by evaluating the practical system function.