Comparing natural image pretraining with digital pathology foundation models for whole slide image-based survival analysis
File(s)
Accepted version
Author(s)
Papadopoulos, Kleanthis Marios
Stathaki, Tania
Barmpoutis, Panagiotis
Benzerdjeb, Nazim
Type
Conference Paper
Abstract
Due to their high information density, Whole Slide Images (WSIs) are considered invaluable tools for survival analysis. Many existing Multiple Instance Learning (MIL) frameworks for this task rely on a ResNet50 backbone pre-trained on natural images. In this work, we leverage recently introduced histopathology foundation models, such as UNI and Hibou, to improve the predictive prowess of existing MIL frameworks. Additionally, we demonstrate that deploying an ensemble of digital pathology foundation models yields higher accuracy, although the gains diminish when applied to more complex MIL architectures.
Editor(s)
Ma, Jixin
Date Issued
2025-09-19
Date Acceptance
2025-06-01
Citation
Sixth International Conference on Computer Vision and Information Technology (CVIT 2025), 2025, 13796, pp.1-12
Publisher
SPIE
Start Page
1
End Page
12
Journal / Book Title
Sixth International Conference on Computer Vision and Information Technology (CVIT 2025)
Volume
13796
License URL
Source
Sixth International Conference on Computer Vision and Information Technology (CVIT 2025)
Publication Status
Published
Start Date
2025-06-20
Finish Date
2025-06-22
Coverage Spatial
Florence, Italy