RETNet: Resolution enhancement Transformer network for magnetic particle imaging based on X-space

Lishuang Guo, Chenbin Ma, Zhen Dong, Jie Tian, Yu An, Jiangang Liu

Abstract: Highlights•RETNet is a feasible, effective and low-cost method for the improvement of MPI resolution.•CST and RST leverage textural and structural features to ensure the clarity and integrity of reconstructed image.•Ablation experiments demonstrate the contribution of the two modules and the importance of the integration of them.•Diverse datasets including simulation, phantom and in vivo datasets, prove the effectiveness and practicability of RETNet.