You need to agree to share your contact information to access this model
This repository is publicly accessible, but you have to accept the conditions to access its files and content.
- This model and associated code are released under the CC-BY-NC-ND 4.0 license and may only be used for non-commercial, academic research purposes with proper attribution.
- Any commercial use, sale, or other monetization of the HistoPLUS model and its derivatives, which include models trained on outputs from the HistoPLUS model or datasets created from the HistoPLUS model, is prohibited and requires prior approval.
- By downloading the model, you attest that all information (affiliation, research use) is correct and up-to-date. Downloading the model requires prior registration on Hugging Face and agreeing to the terms of use. By downloading this model, you agree not to distribute, publish or reproduce a copy of the model. If another user within your organization wishes to use the HistoPLUS model, they must register as an individual user and agree to comply with the terms of use. Users may not attempt to re-identify the deidentified data used to develop the underlying model.
- This model is provided “as-is” without warranties of any kind, express or implied. This model has not been reviewed, certified, or approved by any regulatory body, including but not limited to the FDA (U.S.), EMA (Europe), MHRA (UK), or other medical device authorities. Any application of this model in healthcare or biomedical settings must comply with relevant regulatory requirements and undergo independent validation. Users assume full responsibility for how they use this model and any resulting consequences. The authors, contributors, and distributors disclaim any liability for damages, direct or indirect, resulting from model use. Users are responsible for ensuring compliance with data protection regulations (e.g., GDPR, HIPAA) when using it in research that involves patient data.
Log in or Sign Up to review the conditions and access this model content.
Model Card for HistoPLUS
HistoPLUS addresses critical challenges in analyzing tumor microenvironments (TME) on hematoxylin and eosin (H&E) stained histopathology slides. Existing methods suffer from poor performance on understudied cell types and limited cross-domain generalization, hindering comprehensive TME analysis.
Why it matters:
- Cell detection, segmentation, and classification are fundamental for understanding tumor biology
- Current methods fail on rare cell types not well-represented in public datasets
- Cross-institutional and cross-indication performance remains limited
Our approach: HistoPLUS introduces a CellViT architecture incorporating a state-of-the-art specialized foundation model and trained on a novel and carefully curated pan-cancer dataset of 108,722 nuclei spanning 13 distinct cell types. The model achieves state-of-the-art performance while using significantly fewer parameters, enabling robust analysis of both common and understudied cell populations.
Access Rights
The model is only available to academic and research institutions, for non-commercial use.
Model Description
- Developed by: Owkin, Inc
- Model type: Pretrained model for cell detection, segmentation, and classification (CellViT)
- Training dataset: HistoTRAIN (Proprietary)
- Paper: Arxiv
- License: Owkin non-commercial licence
How To Use (Cell Detection, Segmentation, and Classification)
Python Scripts
The following code snippet allows you to extract cell segmentation masks and to classify cells from histology images using HistoPLUS. These masks and cell types can then be used for downstream analysis, such as nuclei morphometry analysis, cell co-occurrence, density-based biomarker elaboration, etc.
import openslide
from histoplus.extract import extract
from histoplus.helpers.segmentor import CellViTSegmentor
from histoplus.helpers.tissue_detection import detect_tissue_on_wsi
MPP = 0.25 # If available, otherwise set to 0.5
INFERENCE_IMAGE_SIZE = 784
slide = openslide.open_slide("./TCGA-G2-A2EC-01Z-00-DX4.8E4382A4-71F9-4BC3-89AA-09B4F1B54985.svs")
tissue_coords, dz_level = detect_tissue_on_wsi(slide)
segmentor = CellViTSegmentor.from_histoplus(
mpp=MPP,
mixed_precision=True,
inference_image_size=INFERENCE_IMAGE_SIZE,
)
# Process a whole slide image
results = extract(
slide=slide,
coords=tissue_coords,
deepzoom_level=dz_level,
segmentor=segmentor,
batch_size=32,
)
# Save results
results.save("output/results.json")
How To Use (CLI)
The following command allows you to extract cell segmentation masks and to classify cells on entire whole-slide images directly using a command-line interface (CLI).
histoplus \
--slides ./TCGA-G2-A2EC-01Z-00-DX4.8E4382A4-71F9-4BC3-89AA-09B4F1B54985.svs \
--export_dir ./ \
--batch_size 32
Contact
For any additional questions or comments, contact Pierre-Antoine Bannier ([email protected]).
How to cite
@misc{histoplus2025,
title = {Towards Comprehensive Cellular Characterisation of H\&E Slides},
author = {B. Adjadj, P.-A. Bannier, G. Horent, S. Mandela, A. Lyon, K. Schutte, U. Marteau, V. Gaury, L. Dumont, T. Mathieu, R. Belbahri, B. Schmauch, E. Durand, K. Von Loga, L. Gillet},
year = {2025},
eprint = {2508.09926},
archivePrefix= {arXiv},
primaryClass = {cs.CV},
doi = {10.48550/arXiv.2508.09926},
url = {https://arxiv.org/abs/2508.09926}
}
Acknowledgements
The present study was funded by Owkin.
This study makes use of data generated by the MOSAIC consortium (Owkin; Charité – Universitätsmedizin Berlin (DE); Lausanne University Hospital - CHUV (CH); Universitätsklinikum Erlangen (DE); Institut Gustave Roussy (FR); University of Pittsburgh (USA)).
These authors thank Dr Kathrina Alexander, Dr Audrey Caudron, Dr Richard Doughty, Dr Romain Dubois, Dr Thibaut Gioanni, Dr Camelia Radulescu, Dr Thomas Rialland, Dr Pierre Romero and Dr Yannis Roxanis for their contributions to HistoTRAIN and HistoVAL.