Multi-Frame Compensated Super-Resolution for High Spatiotemporal Abdominal 4D-MRI

Motivation: Four-dimensional MRI (4D-MRI) holds significant potential for abdominal radiotherapy, yet it faces a persistent trade-off between spatial and temporal resolution, often leading to undersampled images with motion artifacts. Goal(s): Existing super-resolution models struggle to recover fine details under these conditions. We introduce MCRNet, a multi-frame compensated network designed to enhance abdominal 4D-MRI by leveraging frame redundancy across respiratory cycles. Approach: MCRNet integrates two key modules: the Frame Synergy Attention Module (FSAM) and the Structure-Aware Consolidation Module (SaCM), which together enhance anatomical feature extraction while suppressing artifacts. Results: Comprehensive experiments demonstrate that MCRNet surpasses state-of-the-art methods, effectively restoring anatomical features with minimal artifacts. Impact: MCRNet significantly improves 4D-MRI quality by achieving high spatiotemporal resolution, reducing noise and artifacts, and restoring anatomical structures.