International Journal on Magnetic Particle Imaging IJMPI
Vol. 12 No. 1 Suppl 1 (2026): Int J Mag Part Imag
https://doi.org/10.18416/IJMPI.2026.2603027

Proceedings Articles, ID 974

A Progressive Transformer-GAN Framework for System Matrix Recovery in Magnetic Particle Imaging

Main Article Content

Lizhi Zhang (The Xi’an Key Lab of Radiomics and Intelligent Perception, School of Information Science and Technology, Northwest University, Xi’an, Shaanxi 710069, China), Jintao Li (The Xi’an Key Lab of Radiomics and Intelligent Perception, School of Information Science and Technology, Northwest University, Xi’an, Shaanxi 710069, China), Xiaowei He (The Xi’an Key Lab of Radiomics and Intelligent Perception, School of Information Science and Technology, Northwest University, Xi’an, Shaanxi 710069, China), Hongbo Guo (Northwest University)

Copyright (c) 2026 Lizhi Zhang, Jintao Li, Xiaowei He, Hongbo Guo

Creative Commons License

This work is licensed under a Creative Commons Attribution 4.0 International License.

Abstract

Magnetic Particle Imaging (MPI) is an emerging non-invasive high-resolution technique, but its practicality is constrained by time-consuming repetitive calibration of the System Matrix (SM) when parameters, particle types, or environments change. To address this, we propose TP-GAN, a Transformer-based Progressive GAN for MPI SM super-resolution. It integrates a feature enhancement module to stabilize SM’s physical structure and capture cross-scale correlations, with multi-loss optimization improving consistency between super-resolution and real high-resolution SM, as well as accuracy and anti-noise performance. Experimental results show TP-GAN outperforms existing methods, reducing reliance on repeated calibration and advancing MPI’s biomedical applications.

Article Details