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.
@article{arxiv.2605.25175,
title = {Discrepancy Minimization Improves Cross-Hospital Robustness in Digital Pathology},
author = {Ben Vardi and Dana Schonberger and Yuval Friedmann and Zohar Yakhini and Iris Barshack and Alexander Loebel and Ariel Shamir},
journal= {arXiv preprint arXiv:2605.25175},
year = {2026}
}