Causal Attribution of Model Performance Gaps in Medical Imaging Under Distribution Shifts
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2025
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866912756151615488 |
|---|---|
| author | Gordaliza, Pedro M. Molchanova, Nataliia Banus, Jaume Sanchez, Thomas Cuadra, Meritxell Bach |
| author_facet | Gordaliza, Pedro M. Molchanova, Nataliia Banus, Jaume Sanchez, Thomas Cuadra, Meritxell Bach |
| contents | Deep learning models for medical image segmentation suffer significant performance drops due to distribution shifts, but the causal mechanisms behind these drops remain poorly understood. We extend causal attribution frameworks to high-dimensional segmentation tasks, quantifying how acquisition protocols and annotation variability independently contribute to performance degradation. We model the data-generating process through a causal graph and employ Shapley values to fairly attribute performance changes to individual mechanisms. Our framework addresses unique challenges in medical imaging: high-dimensional outputs, limited samples, and complex mechanism interactions. Validation on multiple sclerosis (MS) lesion segmentation across 4 centers and 7 annotators reveals context-dependent failure modes: annotation protocol shifts dominate when crossing annotators (7.4% $\pm$ 8.9% DSC attribution), while acquisition shifts dominate when crossing imaging centers (6.5% $\pm$ 9.1%). This mechanism-specific quantification enables practitioners to prioritize targeted interventions based on deployment context. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2512_09094 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Causal Attribution of Model Performance Gaps in Medical Imaging Under Distribution Shifts Gordaliza, Pedro M. Molchanova, Nataliia Banus, Jaume Sanchez, Thomas Cuadra, Meritxell Bach Image and Video Processing Computer Vision and Pattern Recognition Machine Learning Methodology Deep learning models for medical image segmentation suffer significant performance drops due to distribution shifts, but the causal mechanisms behind these drops remain poorly understood. We extend causal attribution frameworks to high-dimensional segmentation tasks, quantifying how acquisition protocols and annotation variability independently contribute to performance degradation. We model the data-generating process through a causal graph and employ Shapley values to fairly attribute performance changes to individual mechanisms. Our framework addresses unique challenges in medical imaging: high-dimensional outputs, limited samples, and complex mechanism interactions. Validation on multiple sclerosis (MS) lesion segmentation across 4 centers and 7 annotators reveals context-dependent failure modes: annotation protocol shifts dominate when crossing annotators (7.4% $\pm$ 8.9% DSC attribution), while acquisition shifts dominate when crossing imaging centers (6.5% $\pm$ 9.1%). This mechanism-specific quantification enables practitioners to prioritize targeted interventions based on deployment context. |
| title | Causal Attribution of Model Performance Gaps in Medical Imaging Under Distribution Shifts |
| topic | Image and Video Processing Computer Vision and Pattern Recognition Machine Learning Methodology |
| url | https://arxiv.org/abs/2512.09094 |