Uncertainty Estimation for Pretrained Medical Image Registration Models via Transformation Equivariance
Fuente:
arXiv
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866911412697169920 |
|---|---|
| author | Tian, Lin Hu, Xiaoling Iglesias, Juan Eugenio |
| author_facet | Tian, Lin Hu, Xiaoling Iglesias, Juan Eugenio |
| contents | Accurate image registration is essential in many medical imaging applications, yet most deep registration networks provide little indication of when or where their predictions are unreliable. Existing uncertainty estimation approaches, such as Bayesian methods, ensembles, or MC-dropout, typically require architectural modifications or retraining, precluding their applicability to pretrained registration models. We propose an inference-time, model-agnostic uncertainty estimation framework that applies directly to any pretrained registration network. Our approach is grounded in the transformation equivariance property of image registration, which states that the underlying anatomical mapping should remain consistent under spatial perturbations of the input. Experiments across three pretrained registration models and four anatomical structures show that the resulting uncertainty maps consistently correlate with registration error and highlight unreliably aligned regions. This framework turns pretrained registration networks into risk-aware tools at test time, moving medical image registration closer to safe clinical and large-scale research deployment. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2509_23355 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Uncertainty Estimation for Pretrained Medical Image Registration Models via Transformation Equivariance Tian, Lin Hu, Xiaoling Iglesias, Juan Eugenio Computer Vision and Pattern Recognition Accurate image registration is essential in many medical imaging applications, yet most deep registration networks provide little indication of when or where their predictions are unreliable. Existing uncertainty estimation approaches, such as Bayesian methods, ensembles, or MC-dropout, typically require architectural modifications or retraining, precluding their applicability to pretrained registration models. We propose an inference-time, model-agnostic uncertainty estimation framework that applies directly to any pretrained registration network. Our approach is grounded in the transformation equivariance property of image registration, which states that the underlying anatomical mapping should remain consistent under spatial perturbations of the input. Experiments across three pretrained registration models and four anatomical structures show that the resulting uncertainty maps consistently correlate with registration error and highlight unreliably aligned regions. This framework turns pretrained registration networks into risk-aware tools at test time, moving medical image registration closer to safe clinical and large-scale research deployment. |
| title | Uncertainty Estimation for Pretrained Medical Image Registration Models via Transformation Equivariance |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2509.23355 |