Abstract

The abundance of information present in Whole Slide Images (WSIs) rendersthem an essential tool for survival analysis. Several Multiple InstanceLearning frameworks proposed for this task utilize a ResNet50 backbonepre-trained on natural images. By leveraging recenetly releasedhistopathological foundation models such as UNI and Hibou, the predictiveprowess of existing MIL networks can be enhanced. Furthermore, deploying anensemble of digital pathology foundation models yields higher baselineaccuracy, although the benefits appear to diminish with more complex MILarchitectures. Our code will be made publicly available upon acceptance.

 

Quick Read (beta)