Research map / Image generation / Image restoration
Rethinking Deep Unrolled Model for Accelerated MRI Reconstruction
Bingyu Xin, Meng Ye, Leon Axel, Dimitris N. Metaxas
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 |
|---|---|---|---|
| add AHAxt and AHy together with xt, plus memory st, as CNN inputs (ablation) | Deep unrolled MRI reconstruction · Calgary-Campinas brain 10x | PSNR (dB): +0.36 | Improved |
| add AHAxt or AHy alone to the CNN input (ablation) | Deep unrolled MRI reconstruction · Calgary-Campinas brain 10x | PSNR (dB): +0.00 / +0.01 | No clear effect |
| add learnable memory st to CNN input (ablation) | Deep unrolled MRI reconstruction · Calgary-Campinas brain 10x | PSNR (dB): +0.09 | Improved |
| increase number of previous stages fused in multi-stage feature aggregation (ablation) | Deep unrolled MRI reconstruction · Calgary-Campinas brain 10x | PSNR (dB): 35.88 -> +0.05 (11 stages) | Improved |
| estimate sensitivity maps jointly over adjacent slices (Strategy C) instead of coil-by-coil (A) (ablation) | Sensitivity map estimation · Calgary-Campinas brain 10x | PSNR (dB): 35.60 -> 35.88 | Improved |
| concatenate coils within a slice for SME (Strategy B) (ablation) | Sensitivity map estimation · Calgary-Campinas brain 10x | PSNR (dB): 35.60 -> 35.30 | Hurt |
| adjacent-slice SME on cardiac slices with a large slice gap (ablation) | Sensitivity map estimation · CMRxRecon cardiac 10x | PSNR (dB): 42.75 -> 42.53 | Hurt |
| adjacent SME over time or contrast on cardiac data (ablation) | Sensitivity map estimation · CMRxRecon cardiac 10x | PSNR (dB): 41.67 -> 41.94 (time) | Improved |
| use adjacent-information SME instead of coil-by-coil | Memory consumption · Calgary-Campinas brain | GPU memory: ~55% of previous state-of-the-art | Improved |
Closest papers by meaning
- Time-Embedded Algorithm Unrolling for Computational MRI (NeurIPS 2025 · similarity 0.88)
- A Unified Model for Compressed Sensing MRI Across Undersampling Patterns (CVPR 2025 · similarity 0.87)
- Improving Deep Learning for Accelerated MRI With Data Filtering (NeurIPS 2025 · similarity 0.87)
- Robustness of Deep Learning for Accelerated MRI: Benefits of Diverse Training Data (ICML 2024 · similarity 0.86)
- Efficient Noise Calculation in Deep Learning-based MRI Reconstructions (ICML 2025 · similarity 0.86)
- Progressive Divide-and-Conquer via Subsampling Decomposition for Accelerated MRI (CVPR 2024 · similarity 0.86)
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.
