BCRBench: A Breast Cancer Risk Benchmark

BCRBench is a unified framework for training and benchmarking deep learning models for breast cancer risk prediction from mammography.


🔬 Key Features

  • Multi-view mammography support (e.g., CC, MLO)
  • Longitudinal modeling across multiple screening timepoints
  • Standardized training and evaluation pipeline
  • Modular design for easy integration of new models

📚 Documentation

  • Overview – Scope, objectives, and model comparison
  • Getting Started – Setup and running experiments
  • Datasets – Data formats and preprocessing
  • Models – Architecture details and implementations
  • Training – Training procedures and configuration
  • Evaluation – Metrics and benchmarking

🚀 Quick Start

Follow the Getting Started guide to:

  1. Set up the environment
  2. Prepare datasets
  3. Train and evaluate models

🤝 Contributing

We welcome contributions of new models, datasets, and evaluation methods to extend this benchmark.


📄 License

See repository for licensing details.