Enhancing Brain MRI Super-Resolution Through Multi-Slice Aware Matching and Fusion

Brain MRI super-resolution is a critical task in medical image processing, aiming to enhance the resolution of low-resolution MRI images for better clinical diagnosis and analysis. This paper proposes a Multi-Slice Aware Matching and Fusion (MSAMF) network to enhance brain MRI super-resolution. The network introduces a multi-slice aware module and multi-scale matching strategy to fully utilize multi-slice reference image information, capturing corresponding contextual information from reference features at other scales. Additionally, a multi-scale fusion mechanism is designed to progressively fuse multi-scale matched features, thereby generating more detailed super-resolution images. Experimental results demonstrate the advantages of the network in improving brain MRI reconstruction quality.

Journal article

CAAI Transactions on Intelligence Technology