Investigating the Impact of Histopathological Foundation Models on Regressive Prediction of Homologous Recombination Deficiency

2026-01-29

AI Summary

This study addresses the limited accuracy and generalizability of existing methods in regressing continuous biomarker scores—such as homologous recombination deficiency (HRD) scores—from whole-slide histopathology images. It presents the first systematic evaluation of five state-of-the-art foundation models within a multiple instance learning framework for predicting HRD scores. To mitigate label distribution skew and improve recall for clinically critical but underrepresented patient subgroups, the authors introduce a distribution-aware upsampling strategy. Extensive experiments across breast, endometrial, and lung cancer cohorts demonstrate that the proposed approach significantly outperforms current baselines in prediction accuracy, cross-cohort generalizability, and balanced performance metrics.

Technology Category

• Machine Learning: Multi-instance/Multi-view Learning
• Computer Vision: Other Foundations of Computer Vision
• Humans and AI: Other Foundations of Human Computation & AI

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Foundation models pretrained on large-scale histopathology data have found great success in various fields of computational pathology, but their impact on regressive biomarker prediction remains underexplored. In this work, we systematically evaluate histopathological foundation models for regression-based tasks, demonstrated through the prediction of homologous recombination deficiency (HRD) score - a critical biomarker for personalized cancer treatment. Within multiple instance learning frameworks, we extract patch-level features from whole slide images (WSI) using five state-of-the-art foundation models, and evaluate their impact compared to contrastive learning-based features. Models are trained to predict continuous HRD scores based on these extracted features across breast, endometrial, and lung cancer cohorts from two public medical data collections. Extensive experiments demonstrate that models trained on foundation model features consistently outperform the baseline in terms of predictive accuracy and generalization capabilities while exhibiting systematic differences among the foundation models. Additionally, we propose a distribution-based upsampling strategy to mitigate target imbalance in these datasets, significantly improving the recall and balanced accuracy for underrepresented but clinically important patient populations. Furthermore, we investigate the impact of different sampling strategies and instance bagsizes by ablation studies. Our results highlight the benefits of large-scale histopathological pretraining for more precise and transferable regressive biomarker prediction, showcasing its potential to advance AI-driven precision oncology.

Problem

• homologous recombination deficiency
• regressive prediction
• histopathological foundation models
• biomarker prediction
• whole slide images

Innovation

• histopathological foundation models
• regressive biomarker prediction
• homologous recombination deficiency
• distribution-based upsampling
• whole slide image analysis