Abstract: Spatio-temporal encoding(SPEN) MRI is an ultra-fast magnetic resonance imaging technique. The resolution of the original image acquired with SPEN is relatively low, so super-resolution reconstruction based on sequence physics principles is required to improve the spatial resolution of the original image. Existing SPEN super-resolution reconstruction algorithms based on deep learning have limited ability to capture long-range dependencies. To solve such problems, this paper proposes a transformer-based SPEN MRI super-resolution reconstruction algorithm. An encoder-decoder structure is adopted, and a transformer module is introduced to extract local context information and long-range dependencies of feature maps. Experimental results show that the proposed reconstruction method can reconstruct a super-resolution image with high spatial resolution and no aliasing artifacts from the low-resolution SPEN image without adding additional sampling points. Compared to existing super-resolution methods, the proposed method achieves better results on both clinical and preclinical datasets.

Key words: utrafast MRI, spatio-temporal encoding, deep learning, super-resolution, image reconstruction