Research map / Image generation / Super-resolution
Rethinking Multi-Contrast MRI Super-Resolution: Rectangle-Window Cross-Attention Transformer and Arbitrary-Scale Upsampling
Guangyuan Li, Lei Zhao, Jiakai Sun, Zehua Lan, Zhanjie Zhang, Jiafu Chen, Zhijie Lin, Huaizhong Lin, Wei Xing
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| use rectangle windows (4x16/16x4) instead of 8x8 square windows (ablation) | McASSR · FastMRI 4x | PSNR (dB): 31.63 -> 32.52 | Improved |
| use 4x16 window instead of more elongated 1x64 or 2x32 windows (ablation) | McASSR · FastMRI 4x | PSNR (dB): 31.02 -> 32.52 (1x64); 31.43 -> 32.52 (2x32) | Improved |
| add reference image from another contrast (ablation) | McASSR · FastMRI 4x | PSNR (dB): 31.11 -> 32.52 | Improved |
| add reference reconstruction loss Lref (ablation) | McASSR · FastMRI 4x | PSNR (dB): 32.26 -> 32.52 | Improved |
| add reference-aware term in implicit attention (RIA) (ablation) | McASSR · FastMRI 4x | PSNR (dB): 31.69 -> 32.52 | Improved |
| add grid size to implicit attention (ablation) | McASSR · FastMRI 4x | PSNR (dB): 31.42 -> 32.52 | Improved |
| add periodic (sin) activation in implicit attention (ablation) | McASSR · FastMRI 4x | PSNR (dB): 31.85 -> 32.52 | Improved |
| add cross-attention fusion in RCB (ablation) | McASSR · FastMRI 4x | PSNR (dB): 31.52 -> 32.52 | Improved |
| add Local Feature Enhanced Module instead of single conv (ablation) | McASSR · FastMRI 4x | PSNR (dB): 32.01 -> 32.52 | Improved |
| use implicit upsampling network instead of simplest INR (ablation) | McASSR · FastMRI 8x (out-of-scale) | PSNR (dB): 28.03 -> 28.76 | Improved |
| use McASSR single model versus per-scale fixed-scale multi-contrast methods | McASSR · FastMRI Knee 4x | PSNR (dB): 31.96 -> 32.52 | Improved |
Closest papers by meaning
- Rethinking Diffusion Model for Multi-Contrast MRI Super-Resolution (CVPR 2024 · similarity 0.89)
- MFTN: Multi-Level Feature Transfer Network Based on MRI-Transformer for MR Image Super-resolution (AAAI 2024 · similarity 0.87)
- Boosting ViT-based MRI Reconstruction from the Perspectives of Frequency Modulation, Spatial Purification, and Scale Diversification (AAAI 2025 · similarity 0.86)
- Null Space Matters: Range-Null Decomposition for Consistent Multi-Contrast MRI Reconstruction (AAAI 2024 · similarity 0.86)
- A Unified Model for Compressed Sensing MRI Across Undersampling Patterns (CVPR 2025 · similarity 0.86)
- D2SA: Dual-Stage Distribution and Slice Adaptation for Efficient Test-Time Adaptation in MRI Reconstruction (NeurIPS 2025 · similarity 0.85)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
