Research map / Architectures and efficiency / RNNs and normalization
Deep ADMM-Net for Compressive Sensing MRI
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| train the network instead of using its DCT-based initialization (ablation) | ADMM-Net13 · Brain MRI, 30% | PSNR (dB): 32.71 -> 39.84 | Helped |
| train the network instead of its initialization (ablation) | ADMM-Net13 · Brain MRI, 20% | NMSE: 0.1394 -> 0.0752 | Helped |
| use eight 3x3 DCT filters instead of 1x3 and 3x1 gradient filters (ablation) | ADMM-Net13 · Brain MRI, 20% | NMSE: 0.0899 -> 0.0752 | Helped |
| use eight 3x3 DCT filters instead of 1x3 and 3x1 gradient filters (ablation) | ADMM-Net13 · Brain MRI, 20% | PSNR (dB): 36.52 -> 37.01 | Helped |
| increase filter size to 5x5 with 8 DCT-initialized filters (ablation) | ADMM-Net13 · Brain MRI, 20% | accuracy and time | No clear effect |
| increase number of stages from 13 to 15 (ablation) | ADMM-Net · Brain MRI, 20% | NMSE: 0.0752 -> 0.0739 | Helped |
| increase number of stages beyond 8 (ablation) | ADMM-Net · Brain MRI, 20% | NMSE | No clear effect |
| increase stages from 13 to 14 at 30% sampling (ablation) | ADMM-Net · Brain MRI, 30% | PSNR (dB): 39.70 -> 39.78 | Helped |
| compare ADMM-Net15 with PANO and FDLCP at 30% sampling | ADMM-Net15 · Brain MRI, 30% | PSNR (dB): 39.13 -> 39.84 | Helped |
| compare ADMM-Net15 with BM3D-MRI at 30% sampling | ADMM-Net15 · Brain MRI, 30% | PSNR (dB): 40.33 -> 39.84 | Hurt |
| apply net trained on brain data to chest data | ADMM-Net15-B · Chest MRI, 20% | PSNR (dB): 37.68 vs FDLCP 37.77 | Mixed |
| train on chest data instead of brain data (ablation) | ADMM-Net15 · Chest MRI, 20% | NMSE: 0.0790 -> 0.0775 | Helped |
Closest papers by meaning
- A Unified Model for Compressed Sensing MRI Across Undersampling Patterns (CVPR 2025 · similarity 0.89)
- UFC-Net: Unrolling Fixed-point Continuous Network for Deep Compressive Sensing (CVPR 2024 · similarity 0.85)
- Optimizing ADMM and Over-Relaxed ADMM Parameters for Linear Quadratic Problems (AAAI 2024 · similarity 0.85)
- Multi-Cross Sampling and Frequency-Division Reconstruction for Image Compressed Sensing (AAAI 2024 · similarity 0.84)
- Progressive Divide-and-Conquer via Subsampling Decomposition for Accelerated MRI (CVPR 2024 · similarity 0.84)
- Fast MRI for All: Bridging Access Gaps by Training without Raw Data (NeurIPS 2025 · similarity 0.84)
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.