klotz11/plism-dataset-tiles-st

Stain-Transferred PLISM Dataset

Dataset Overview

The Stain-Transferred PLISM dataset is a synthetic variant of the PLISM dataset tiles provided by Filiot et al. (2025), which is based on the original PLISM-wsi dataset by Ochi et al. (2024).

This dataset isolates global color-level variations (staining profiles) from localized morphological and scanner-specific hardware artifacts. It is specifically designed to evaluate and robustify computational pathology foundation models against purely color-driven domain shifts. Dataset Creation

To isolate color variations, we modified the Owkin PLISM tiles dataset through the following pipeline:

• —Anchor Slide: We selected a single slide (GIVH_AT2_to_GMH_S60) to serve as the morphological anchor.
• —Color Transfer: We generated synthetic variants by transferring the color profiles of all other target slides within the cohort onto this anchor. A Reinhard stain normalizer was fitted to a composite target image (25 random tissue patches from the target slide) and applied to all patches of the anchor slide.
• —Result: A dataset where tissue morphology and spatial structures are constant across all samples, leaving high-level color statistics as the sole source of variance.

Acknowledgments

We thank the authors of the original PLISM dataset (Ochi et al., 2024) for their contribution, and the team at Owkin (Filiot et al., 2025) for providing the highly accessible tile extractions that served as the foundation for this dataset.

License

This dataset is distributed under the CC BY 4.0 license, inheriting the licensing of the original PLISM dataset and the Owkin tile extractions.

How to Cite

If you use this dataset, please cite the original PLISM authors:

Owkin PLISM Tiles

@inproceedings{filiot2025distilling,
  title={Distilling foundation models for robust and efficient models in digital pathology},
  author={Filiot, Alexandre and Dop, Nicolas and Tchita, Oussama and Riou, Auriane and Dubois, R{\'e}my and Peeters, Thomas and Valter, Daria and Scalbert, Marin and Saillard, Charlie and Robin, Genevi{\`e}ve and others},
  booktitle={International Conference on Medical Image Computing and Computer-Assisted Intervention},
  pages={162--172},
  year={2025},
  organization={Springer}
}

Original PLISM Dataset

@article{ochi2024registered,
  title={Registered multi-device/staining histology image dataset for domain-agnostic machine learning models},
  author={Ochi, Mieko and Komura, Daisuke and Onoyama, Takumi and Shinbo, Koki and Endo, Haruya and Odaka, Hiroto and Kakiuchi, Miwako and Katoh, Hiroto and Ushiku, Tetsuo and Ishikawa, Shumpei},
  journal={Scientific Data},
  volume={11},
  number={1},
  pages={330},
  year={2024},
  publisher={Nature Publishing Group UK London}
}