Ensemble learning of foundation models for precision oncology

2025-08-22

AI Summary

Current pathology foundation models suffer from weak generalizability and inconsistent performance due to heterogeneous training strategies and datasets, hindering unified clinical deployment in precision oncology. To address this, we propose ELF—a novel ensemble framework for whole-slide images (WSIs)—the first multi-foundation-model integration method tailored to WSI analysis. ELF fuses representations from five state-of-the-art models to generate robust, unified WSI-level embeddings. By aggregating complementary cross-model information and modeling features at the sliding-window level, ELF overcomes contextual fragmentation inherent in conventional tile-level analysis. Evaluated across 20 anatomical sites and over 50,000 WSIs, ELF significantly outperforms all individual foundation models and existing sliding-level approaches on disease classification, biomarker detection, and treatment response prediction—particularly demonstrating superior generalizability in data-scarce settings.

Technology Category

• Machine Learning: Other Foundations of Machine Learning
• Computer Vision: Other Foundations of Computer Vision
• Multiagent Systems: Other Foundations of Multi Agent Systems

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
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

Histopathology is essential for disease diagnosis and treatment decision-making. Recent advances in artificial intelligence (AI) have enabled the development of pathology foundation models that learn rich visual representations from large-scale whole-slide images (WSIs). However, existing models are often trained on disparate datasets using varying strategies, leading to inconsistent performance and limited generalizability. Here, we introduce ELF (Ensemble Learning of Foundation models), a novel framework that integrates five state-of-the-art pathology foundation models to generate unified slide-level representations. Trained on 53,699 WSIs spanning 20 anatomical sites, ELF leverages ensemble learning to capture complementary information from diverse models while maintaining high data efficiency. Unlike traditional tile-level models, ELF's slide-level architecture is particularly advantageous in clinical contexts where data are limited, such as therapeutic response prediction. We evaluated ELF across a wide range of clinical applications, including disease classification, biomarker detection, and response prediction to major anticancer therapies, cytotoxic chemotherapy, targeted therapy, and immunotherapy, across multiple cancer types. ELF consistently outperformed all constituent foundation models and existing slide-level models, demonstrating superior accuracy and robustness. Our results highlight the power of ensemble learning for pathology foundation models and suggest ELF as a scalable and generalizable solution for advancing AI-assisted precision oncology.

Problem

• Integrating disparate pathology foundation models for consistent performance
• Improving generalizability of AI models across diverse cancer types
• Enhancing clinical decision-making with limited histopathology data

Innovation

• Ensemble learning integrates multiple pathology foundation models
• Generates unified slide-level representations from diverse data
• Superior accuracy in clinical oncology applications