Research map / Image generation / Image restoration
Adaptive Correspondence Scoring for Unsupervised Medical Image Registration
Xiaoran Zhang, John C. Stendahl, Lawrence H. Staib, Albert J. Sinusas, Alex Wong, James S. Duncan
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 correspondence scoring map to re-weight reconstruction residuals | Voxelmorph, Transmorph, Diffusemorph · ACDC | DSC (%): 79.48 -> 80.50, 76.94 -> 78.39, 67.38 -> 72.09 | Improved |
| add correspondence scoring map to re-weight reconstruction residuals | Voxelmorph, Transmorph, Diffusemorph · CAMUS | DSC (%): 81.50 -> 81.74, 79.24 -> 79.64, 75.23 -> 77.65 | Improved |
| add aleatoric uncertainty weighting (NLL, beta-NLL) | Voxelmorph, Transmorph, Diffusemorph · ACDC and CAMUS | DSC (%): 79.48 -> 76.49 (NLL on Voxelmorph, ACDC) | Hurt |
| add adaptive weighting schemes (AdaReg, AdaFrame) | Voxelmorph, Transmorph, Diffusemorph · ACDC | DSC (%): 79.48 -> 66.38 (AdaFrame on Voxelmorph) | Hurt |
| add momentum-guided smoothness regularization on the scoring map (ablation) | AdaCS with Diffusemorph · ACDC | DSC (%): 71.62 -> 72.09 | Improved |
| add momentum-guided smoothness regularization on the scoring map (ablation) | AdaCS with Transmorph · ACDC | DSC (%): 77.84 -> 78.39 | Improved |
| add momentum-guided smoothness regularization on the scoring map (ablation) | AdaCS with Voxelmorph · CAMUS | DSC (%): 81.58 -> 81.74 | No clear effect |
| remove the regularization L_reg on the scoring map (ablation) | AdaCS scoring estimator · ACDC and CAMUS | scoring map | Hurt |
| score from target image only instead of target and reconstructed target (ablation) | AdaCS with Diffusemorph · ACDC | DSC (%): 69.05 -> 72.09 | Improved |
| add AdaCS to Transmorph on private 3D echo | Transmorph · Private 3D Echo | DSC (%): 74.89 -> 75.54 | No clear effect |
| add AdaCS to Voxelmorph and cLapIRN on OASIS | Voxelmorph, cLapIRN · OASIS 3D (average Dice) | DSC (%): 67.80 -> 69.32, 76.41 -> 76.87 | Improved |
| replace MSE with robust losses (NCC, MI, MAE) without AdaCS (ablation) | Voxelmorph · ACDC | DSC (%): 79.48 -> 78.55 (NCC), 78.04 (MI), 78.27 (MAE) | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- Diffusemorph — DiffuseMorph: Unsupervised Deformable Image Registration Using Diffusion Model (ECCV 2022) · better on 2 of 2 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Deep Learning in Medical Image Registration: Magic or Mirage? (NeurIPS 2024 · similarity 0.87)
- Uncertainty Estimation for Pretrained Medical Image Registration Models via Transformation Equivariance (NeurIPS 2026 · similarity 0.86)
- CARL: A Framework for Equivariant Image Registration (CVPR 2025 · similarity 0.85)
- Correlation-aware Coarse-to-fine MLPs for Deformable Medical Image Registration (CVPR 2024 · similarity 0.84)
- Adaptive Bidirectional Displacement for Semi-Supervised Medical Image Segmentation (CVPR 2024 · similarity 0.84)
- Leveraging Anatomical Consistency for Multi-Object Detection in Ultrasound Images via Source-free Unsupervised Domain Adaptation (AAAI 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.
