Research map / Image generation / Fast diffusion sampling
DiffuseMorph: Unsupervised Deformable Image Registration Using Diffusion Model
Boah Kim, Inhwa Han, Jong Chul Ye
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 |
|---|---|---|---|
| increase registration loss weight lambda from 2 to 10 (ablation) | DiffuseMorph · OASIS-3 brain MR registration | Dice: 0.736 -> 0.756 | Improved |
| increase registration loss weight lambda from 2 to 10 (ablation) | DiffuseMorph · OASIS-3 brain MR registration | non-positive Jacobian (%): 0.498 -> 0.505 | Hurt |
| use diffusion-based latent features instead of VM in facial registration | VoxelMorph deformation network · RaFD facial expression registration | NMSE x10^-1: 0.047 -> 0.032 | Improved |
| use DiffuseMorph instead of VM-diff in facial registration | VM-diff · RaFD facial expression registration | SSIM: 0.957 -> 0.964 | Improved |
| use DiffuseMorph instead of VM-diff on folding | VM-diff · RaFD facial expression registration | non-positive Jacobian (%): 0.014 -> 0.017 | No clear effect |
| use DiffuseMorph instead of VM in cardiac registration | VM · ACDC cardiac MR registration | Dice: 0.787 -> 0.802 | Improved |
| use DiffuseMorph instead of VM-diff in cardiac registration | VM-diff · ACDC cardiac MR registration | non-positive Jacobian (%): 0.291 -> 0.161 | Improved |
| scale the latent score feature instead of scaling or integrating the field for continuous deformation | VM-diff · ACDC cardiac MR continuous deformation | PSNR (dB): 29.481 -> 30.725 | Improved |
| scale the latent score feature instead of scaling the field | VM · ACDC cardiac MR continuous deformation | PSNR (dB): 30.678 -> 30.725 | No clear effect |
| use DiffuseMorph instead of VM in brain registration | VM · OASIS-3 brain MR registration | Dice: 0.749 -> 0.756 | Improved |
| use DiffuseMorph instead of SyN in brain registration | SyN · OASIS-3 brain MR registration | Dice: 0.752 -> 0.756 | Improved |
| use DiffuseMorph instead of MSDIRNet | MSDIRNet · OASIS-3 brain MR registration | non-positive Jacobian (%): 0.804 -> 0.505 | Improved |
Later papers that compared with this method
2 later accepted papers report DiffuseMorph as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- Learning Patient-Specific Disease Dynamics With Latent Flow Matching For Longitudinal Imaging Generation (ICLR 2026 · better on 0 of 1 shared result)
- Adaptive Correspondence Scoring for Unsupervised Medical Image Registration (ECCV 2024 · 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
- NODEO: A Neural Ordinary Differential Equation Based Optimization Framework for Deformable Image Registration (CVPR 2022 · similarity 0.89)
- An Unsupervised Learning Model for Deformable Medical Image Registration (CVPR 2018 · similarity 0.89)
- A Plug-and-Play Image Registration Network (ICLR 2024 · similarity 0.86)
- DiffusionBlend: Learning 3D Image Prior through Position-aware Diffusion Score Blending for 3D Computed Tomography Reconstruction (NeurIPS 2024 · similarity 0.85)
- As-Plausible-As-Possible: Plausibility-Aware Mesh Deformation Using 2D Diffusion Priors (CVPR 2024 · similarity 0.84)
- Correlation-aware Coarse-to-fine MLPs for Deformable Medical Image Registration (CVPR 2024 · 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.
