DINOv3 with Test-Time Training for Medical Image Registration
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866915453015687168 |
|---|---|
| author | Wang, Shansong Safari, Mojtaba Hu, Mingzhe Li, Qiang Chang, Chih-Wei Qiu, Richard LJ Yang, Xiaofeng |
| author_facet | Wang, Shansong Safari, Mojtaba Hu, Mingzhe Li, Qiang Chang, Chih-Wei Qiu, Richard LJ Yang, Xiaofeng |
| contents | Prior medical image registration approaches, particularly learning-based methods, often require large amounts of training data, which constrains clinical adoption. To overcome this limitation, we propose a training-free pipeline that relies on a frozen DINOv3 encoder and test-time optimization of the deformation field in feature space. Across two representative benchmarks, the method is accurate and yields regular deformations. On Abdomen MR-CT, it attained the best mean Dice score (DSC) of 0.790 together with the lowest 95th percentile Hausdorff Distance (HD95) of 4.9+-5.0 and the lowest standard deviation of Log-Jacobian (SDLogJ) of 0.08+-0.02. On ACDC cardiac MRI, it improves mean DSC to 0.769 and reduces SDLogJ to 0.11 and HD95 to 4.8, a marked gain over the initial alignment. The results indicate that operating in a compact foundation feature space at test time offers a practical and general solution for clinical registration without additional training. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2508_14809 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | DINOv3 with Test-Time Training for Medical Image Registration Wang, Shansong Safari, Mojtaba Hu, Mingzhe Li, Qiang Chang, Chih-Wei Qiu, Richard LJ Yang, Xiaofeng Computer Vision and Pattern Recognition Artificial Intelligence Prior medical image registration approaches, particularly learning-based methods, often require large amounts of training data, which constrains clinical adoption. To overcome this limitation, we propose a training-free pipeline that relies on a frozen DINOv3 encoder and test-time optimization of the deformation field in feature space. Across two representative benchmarks, the method is accurate and yields regular deformations. On Abdomen MR-CT, it attained the best mean Dice score (DSC) of 0.790 together with the lowest 95th percentile Hausdorff Distance (HD95) of 4.9+-5.0 and the lowest standard deviation of Log-Jacobian (SDLogJ) of 0.08+-0.02. On ACDC cardiac MRI, it improves mean DSC to 0.769 and reduces SDLogJ to 0.11 and HD95 to 4.8, a marked gain over the initial alignment. The results indicate that operating in a compact foundation feature space at test time offers a practical and general solution for clinical registration without additional training. |
| title | DINOv3 with Test-Time Training for Medical Image Registration |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence |
| url | https://arxiv.org/abs/2508.14809 |