This dataset contains annotated 3D mammography (Digital Breast Tomosynthesis, DBT) imaging data for research in breast cancer detection using machine learning.

The dataset was developed through a collaboration between iMerit, Segmed, and Advocate Health.

Full dataset access:

https://imerit.net/3d-mammogram-dataset/

Official announcement:

https://imerit.net/press/imerit-segmed-and-advocate-health-announce-free-annotated-3d-mammography-data-set-to-advance-ai-research-in-breast-cancer-detection/

Dataset

The dataset contains:

3D mammography (DBT) imaging studies
Expert annotations for breast cancer detection
Associated study metadata

Imaging data is provided in standard medical imaging formats (e.g., DICOM) with structured annotations.

Example Data

The following figure shows example slices from a Digital Breast Tomosynthesis (DBT) study along with corresponding annotations.

Example dataset structure

sample_data/
├── samplecase1/
│   ├── *.dcm
│   ├── *.dcm
│   └── ...
│
├── annotations.json
│
└── metadata.xlsx

Access

The full dataset can be downloaded from:

https://imerit.net/3d-mammogram-dataset/

License

This dataset is released under the Segmed–iMerit Joint Research Use Agreement.

https://www.segmed.ai/legal/joint-research-use-agreement

Use is restricted to non-commercial research purposes.

Citation

iMerit, Segmed, and Advocate Health.
3D Mammography (Digital Breast Tomosynthesis) Dataset.
DOI: https://doi.org/10.64898/2026.03.03.25337924

About iMerit

iMerit provides high-quality data annotation and AI training data solutions, helping organizations build machine learning systems across healthcare, medical imaging, autonomous mobility, and other AI-driven domains.

Downloads last month: 35