Comparing ImageNet Pre-training with Digital Pathology Foundation Models for Whole Slide Image-Based Survival Analysis

TL;DR

This paper evaluates the effectiveness of digital pathology foundation models like UNI and Hibou in improving Whole Slide Image-based survival analysis, comparing them to traditional ImageNet pre-trained models within MIL frameworks.

Contribution

It introduces the use of recent histopathological foundation models for survival analysis and assesses their impact on MIL network performance.

Findings

Foundation models improve baseline accuracy

Ensemble models yield higher baseline performance

Benefits diminish with complex MIL architectures

Abstract

The abundance of information present in Whole Slide Images (WSIs) renders them an essential tool for survival analysis. Several Multiple Instance Learning frameworks proposed for this task utilize a ResNet50 backbone pre-trained on natural images. By leveraging recenetly released histopathological foundation models such as UNI and Hibou, the predictive prowess of existing MIL networks can be enhanced. Furthermore, deploying an ensemble of digital pathology foundation models yields higher baseline accuracy, although the benefits appear to diminish with more complex MIL architectures. Our code will be made publicly available upon acceptance.