SPIDER Thorax Model

🚀 SPIDER-Thorax Model

The SPIDER-thorax model is a deep - learning model designed for patch - level pathology classification of the thorax. It is part of the SPIDER dataset initiative, which offers a large - scale, high - quality, multi - organ pathology dataset with expert - annotated labels. If you're interested in supporting, sponsoring, or obtaining a commercial license for the SPIDER data and models, reach out to us at models@hist.ai.

🚀 Quick Start

The model operates on 1120×1120 patches. Use the following code to load and utilize the model:

Basic Usage

from transformers import AutoModel, AutoProcessor
from PIL import Image

model = AutoModel.from_pretrained("histai/SPIDER-thorax-model", trust_remote_code=True)
processor = AutoProcessor.from_pretrained("histai/SPIDER-thorax-model", trust_remote_code=True)

image = Image.open("path_to_image.png")
inputs = processor(images=image, return_tensors="pt")
outputs = model(**inputs)
print(outputs.predicted_class_names)

✨ Features

• Specialized for patch - level pathology classification of the thorax.
• Part of the comprehensive SPIDER dataset initiative.

📦 Installation

The installation is mainly about using the transformers library. If you haven't installed it yet, you can use the following command:

pip install transformers

📚 Documentation

Model Sources

• Repository: https://github.com/HistAI/SPIDER
• Paper: SPIDER: A Comprehensive Multi-Organ Supervised Pathology Dataset and Baseline Models

Training Data

The model is trained on the [SPIDER - thorax](https://huggingface.co/datasets/histai/SPIDER - thorax) dataset, a subset of the SPIDER dataset. The dataset details are as follows:

The dataset includes:

Total Counts:

• 78,307 central patches
• 599,459 total patches (including context patches)
• 411 total slides used for annotation

Results

🔧 Technical Details

The model is designed for patch - level pathology classification of the thorax. It is trained on a specific subset of the SPIDER dataset, which provides a rich set of expert - annotated data for different classes related to the thorax. The model works with a specific patch size of 1120×1120 and uses the transformers library for loading and processing.

📄 License

The model is licensed under CC BY - NC 4.0 and is for research use only.

Citation

If you use this model, please cite the following:

@misc{nechaev2025spidercomprehensivemultiorgansupervised,
      title={SPIDER: A Comprehensive Multi-Organ Supervised Pathology Dataset and Baseline Models}, 
      author={Dmitry Nechaev and Alexey Pchelnikov and Ekaterina Ivanova},
      year={2025},
      eprint={2503.02876},
      archivePrefix={arXiv},
      primaryClass={eess.IV},
      url={https://arxiv.org/abs/2503.02876}, 
}

More Information

To explore other models and the SPIDER dataset, you can visit the Hugging Face HistAI page or GitHub of the project.

Contacts

• Authors: Dmitry Nechaev, Alexey Pchelnikov, Ekaterina Ivanova
• Email: dmitry@hist.ai, alex@hist.ai, kate@hist.ai