A deep learning framework for efficient pathology image analysis

TL;DR

EAGLE is a deep learning framework that significantly improves efficiency and accuracy in pathology image analysis by selectively examining informative regions, enabling real-time processing and broader accessibility.

Contribution

The paper introduces EAGLE, a novel selective analysis framework that reduces computational costs and enhances performance in pathology image tasks compared to existing methods.

Findings

Outperforms state-of-the-art methods by up to 23% in AUROC

Reduces processing time by over 99%

Enables real-time, accessible pathology analysis

Abstract

Artificial intelligence (AI) has transformed digital pathology by enabling biomarker prediction from high-resolution whole-slide images (WSIs). However, current methods are computationally inefficient, processing thousands of redundant tiles per WSI and requiring complex aggregator models. We introduce EAGLE (Efficient Approach for Guided Local Examination), a deep learning framework that emulates pathologists by selectively analyzing informative regions. EAGLE incorporates two foundation models: CHIEF for efficient tile selection and Virchow2 for extracting high-quality features. Benchmarking was conducted against leading slide- and tile-level foundation models across 43 tasks from nine cancer types, spanning morphology, biomarker prediction, treatment response and prognosis. EAGLE outperformed state-of-the-art patch aggregation methods by up to 23% and achieved the highest AUROC overall. It processed a slide in 2.27 seconds, reducing computational time by more than 99% compared to existing models. This efficiency enables real-time workflows, allows rapid review of the exact tiles used for each prediction, and reduces dependence on high-performance computing, making AI-powered pathology more accessible. By reliably identifying meaningful regions and minimizing artifacts, EAGLE provides robust and auditable outputs, supported by systematic negative controls and attention concentration analyses. Its unified embedding enables rapid slide searches, integration into multi-omics pipelines and emerging clinical foundation models.