Diversity Over Scale: Whole-Slide Image Variety Enables H&E Foundation Model Training with Fewer Patches
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866917078641934336 |
|---|---|
| author | Bosch, Christoph Wong, John K. L. Paulikat, Martin Zapukhlyak, Myroslav Arora, Bharti Aichmüller-Ratnaparkhe, Manasi Baumann, Jens Karn, Shivani Kamble, Rutuja Karnik, Swapnil Khedkar, Bhushan Chhut, Serey Vathana Aswolinskiy, Witali Aichmüller, Christian |
| author_facet | Bosch, Christoph Wong, John K. L. Paulikat, Martin Zapukhlyak, Myroslav Arora, Bharti Aichmüller-Ratnaparkhe, Manasi Baumann, Jens Karn, Shivani Kamble, Rutuja Karnik, Swapnil Khedkar, Bhushan Chhut, Serey Vathana Aswolinskiy, Witali Aichmüller, Christian |
| contents | Rapid progress in computational pathology is increasingly driven by vision foundation models pretrained on vast histopathology datasets. While recent efforts have prioritized training on an ever-larger amount of patches, we take an alternative approach focused on data diversity. Our foundation model, Athena, was initialized from a pretrained model and trained on just 115 million tissue patches, several times fewer than recent histopathology foundation models. Rather than relying on patch volume or complex sampling heuristics, we maximize data diversity by randomly selecting only a moderate number of patches per whole-slide image from our diverse internal repository, which spans multiple countries, institutions, and scanner types. Evaluated on a single patch-level benchmark and four slide-level downstream tasks (two molecular and two morphological), Athena approaches the state-of-the-art and even surpasses several models trained on substantially larger datasets. This indicates that diversity across whole-slide images, rather than patch quantity alone, drives learning in histopathology foundation models. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2511_10286 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Diversity Over Scale: Whole-Slide Image Variety Enables H&E Foundation Model Training with Fewer Patches Bosch, Christoph Wong, John K. L. Paulikat, Martin Zapukhlyak, Myroslav Arora, Bharti Aichmüller-Ratnaparkhe, Manasi Baumann, Jens Karn, Shivani Kamble, Rutuja Karnik, Swapnil Khedkar, Bhushan Chhut, Serey Vathana Aswolinskiy, Witali Aichmüller, Christian Tissues and Organs Rapid progress in computational pathology is increasingly driven by vision foundation models pretrained on vast histopathology datasets. While recent efforts have prioritized training on an ever-larger amount of patches, we take an alternative approach focused on data diversity. Our foundation model, Athena, was initialized from a pretrained model and trained on just 115 million tissue patches, several times fewer than recent histopathology foundation models. Rather than relying on patch volume or complex sampling heuristics, we maximize data diversity by randomly selecting only a moderate number of patches per whole-slide image from our diverse internal repository, which spans multiple countries, institutions, and scanner types. Evaluated on a single patch-level benchmark and four slide-level downstream tasks (two molecular and two morphological), Athena approaches the state-of-the-art and even surpasses several models trained on substantially larger datasets. This indicates that diversity across whole-slide images, rather than patch quantity alone, drives learning in histopathology foundation models. |
| title | Diversity Over Scale: Whole-Slide Image Variety Enables H&E Foundation Model Training with Fewer Patches |
| topic | Tissues and Organs |
| url | https://arxiv.org/abs/2511.10286 |