Discrepancy Minimization Improves Cross-Hospital Robustness in Digital Pathology
About
Pathology foundation models (PFMs) have advanced rapidly in recent years and support training classifiers for a range of histopathology tasks. However, their robustness across hospitals remains limited: performance often degrades when training a classifier on data from one hospital and evaluating it on another target hospital. We address this challenge by fine-tuning PFMs with a local maximum mean discrepancy (LMMD) objective that applies to two settings: domain adaptation, where unlabeled target-hospital data is available, and domain generalization, where target-hospital data is unavailable at all. Experiments at both the patch- and slide-level show consistent improvements across multiple PFMs and tasks.
Ben Vardi, Dana Schonberger, Yuval Friedmann, Zohar Yakhini, Iris Barshack, Alexander Loebel, Ariel Shamir• 2026
Related benchmarks
| Task | Dataset | Result | Rank | |
|---|---|---|---|---|
| Slide-level classification | TCGA NSCLC (test) | -- | 96 | |
| Domain Adaptation | NSCLC patch-level | Balanced Accuracy84.88 | 12 | |
| Domain Adaptation | RCC patch-level | Balanced Accuracy96.85 | 10 | |
| Patch-Level Classification | NSCLC Unseen target hospitals | Balanced Accuracy83.79 | 10 | |
| Slide-level classification | RCC Slide-level (test) | Balanced Accuracy97.17 | 10 | |
| Domain Adaptation | RCC patch-level (Gen) | Balanced Accuracy94.1 | 2 | |
| Domain Adaptation | NSCLC patch-level (Unbalanced) | Balanced Accuracy80.48 | 2 | |
| Domain Adaptation | RCC patch-level (Unbalanced) | Balanced Accuracy94.4 | 2 |
Showing 8 of 8 rows