Research map / Image generation / Image restoration
Fast Symmetric Diffeomorphic Image Registration with Convolutional Neural Networks
Check 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 |
|---|---|---|---|
| Use SYMNet instead of the strongest VM baseline (λ=1) | OASIS atlas-based registration · OASIS atlas-based registration | Avg. DSC: 0.727 (0.144) -> 0.738 (0.108) | Helped |
| Use SYMNet instead of SyN | OASIS atlas-based registration · OASIS atlas-based registration | Runtime (s): 1039 -> 0.414 | Helped |
| Add local orientation-consistency loss (λ₁=1000) (ablation) | SYMNet · OASIS atlas-based registration | Non-positive Jacobian voxels (|Jφ|≤0): 1156 (2015) -> 0.471 (0.921) | Helped |
| Add local orientation-consistency loss (λ₁=1000) (ablation) | SYMNet · OASIS atlas-based registration | Avg. DSC: 0.743 (0.113) -> 0.738 (0.108) | Hurt |
| Set orientation-consistency weight λ₁ from 0 to 1 (ablation) | SYMNet · OASIS atlas-based registration | Non-positive Jacobian voxels (|Jφ|≤0): 1156 (2015) -> 860 (1562) | Helped |
| Set orientation-consistency weight λ₁ from 1 to 10 (ablation) | SYMNet · OASIS atlas-based registration | Non-positive Jacobian voxels (|Jφ|≤0): 860 (1562) -> 460 (845) | Helped |
| Set orientation-consistency weight λ₁ from 10 to 100 (ablation) | SYMNet · OASIS atlas-based registration | Non-positive Jacobian voxels (|Jφ|≤0): 460 (845) -> 133 (260) | Helped |
| Set orientation-consistency weight λ₁ from 100 to 1000 (ablation) | SYMNet · OASIS atlas-based registration | Non-positive Jacobian voxels (|Jφ|≤0): 133 (260) -> 0.471 (0.921) | Helped |
| Increase orientation-consistency weight λ₁ from 0 to 1000 (ablation) | SYMNet · OASIS atlas-based registration | Avg. DSC: 0.7434 (0.113) -> 0.7381 (0.108) | Hurt |
| Set orientation-consistency weight λ₁ from 0 to 1 (ablation) | SYMNet · OASIS atlas-based registration | Avg. DSC: 0.7434 (0.113) -> 0.7431 (0.110) | No clear effect |
| Set orientation-consistency weight λ₁ from 1 to 10 (ablation) | SYMNet · OASIS atlas-based registration | Avg. DSC: 0.7431 (0.110) -> 0.7423 (0.111) | No clear effect |
| Set orientation-consistency weight λ₁ from 10 to 100 (ablation) | SYMNet · OASIS atlas-based registration | Avg. DSC: 0.7423 (0.111) -> 0.7408 (0.104) | No clear effect |
Closest papers by meaning
- NODEO: A Neural Ordinary Differential Equation Based Optimization Framework for Deformable Image Registration (CVPR 2022 · similarity 0.92)
- A Plug-and-Play Image Registration Network (ICLR 2024 · similarity 0.89)
- CARL: A Framework for Equivariant Image Registration (CVPR 2025 · similarity 0.87)
- SACB-Net: Spatial-awareness Convolutions for Medical Image Registration (CVPR 2025 · similarity 0.86)
- Deep Learning in Medical Image Registration: Magic or Mirage? (NeurIPS 2024 · similarity 0.86)
- Correlation-aware Coarse-to-fine MLPs for Deformable Medical Image Registration (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.