Enabling clinical use of foundation models in histopathology

2026-02-25

AI Summary

This study addresses the vulnerability of foundation models in computational pathology to pre-analytical and technical variations—such as tissue processing protocols and scanner differences—which introduce prediction bias in downstream tasks and hinder clinical deployment. To mitigate this, the authors propose a novel robustness-oriented loss function that, without retraining the foundation model, enhances downstream training by steering the model toward biologically relevant features while reducing sensitivity to technical artifacts. The approach is rigorously evaluated across eight state-of-the-art histopathology foundation models using 27,042 whole-slide images from 6,155 patients, with thousands of downstream models systematically trained and assessed. Results demonstrate substantial improvements in generalization and predictive accuracy on real-world data, offering an efficient and practical pathway toward clinically viable computational pathology systems.

Technology Category

• Computer Vision: Other Foundations of Computer Vision
• Machine Learning: Deep Neural Architectures and Foundation Models
• Humans and AI: Other Foundations of Human Computation & AI

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Web Mining and Content Analysis: Large pretrained models with web data
• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems

Abstract

Foundation models in histopathology are expected to facilitate the development of high-performing and generalisable deep learning systems. However, current models capture not only biologically relevant features, but also pre-analytic and scanner-specific variation that bias the predictions of task-specific models trained from the foundation model features. Here we show that introducing novel robustness losses during training of downstream task-specific models reduces sensitivity to technical variability. A purpose-designed comprehensive experimentation setup with 27,042 WSIs from 6155 patients is used to train thousands of models from the features of eight popular foundation models for computational pathology. In addition to a substantial improvement in robustness, we observe that prediction accuracy improves by focusing on biologically relevant features. Our approach successfully mitigates robustness issues of foundation models for computational pathology without retraining the foundation models themselves, enabling development of robust computational pathology models applicable to real-world data in routine clinical practice.

Problem

• foundation models
• histopathology
• technical variability
• robustness
• computational pathology

Innovation

• foundation models
• histopathology
• robustness loss
• computational pathology
• technical variability