BreastDCEDL: A Deep Learning–Ready Breast DCE-MRI Dataset

BreastDCEDL: A Dynamic Contrast-Enhanced Breast MRI Dataset

BreastDCEDL is a curated, deep learning–ready dataset of 3D DCE-MRI scans from 2,070 breast cancer patients across three major clinical trials (I-SPY1, I-SPY2, Duke) available through The Cancer Imaging Archive (TCIA). The dataset includes pre-treatment MRI volumes, 3D tumor segmentations, and harmonized clinical and demographic metadata, including pathologic complete response (pCR), hormone receptor (HR), and HER2 status. This resource supports advanced AI research in breast cancer, enabling training and evaluation of transformer-based and other deep learning models.

2070 patients from TCIA clinical trials
3D MRI scans · Tumor segmentations · Clinical metadata
Labels: pCR, HR, HER2, age, race

Keywords: breast cancer, DCE-MRI, tumor segmentation, deep learning, clinical metadata, ViT, pCR prediction, pCR,HR,HER2,YCIA,ISPY1,DUKE,ISPY2

License info

This dataset is released under CC BY-NC 4.0 to comply with the licensing terms of the source data from The Cancer Imaging Archive (TCIA). The I-SPY 1 dataset is licensed under CC BY 3.0 and I-SPY 2 under CC BY 4.0, both of which permit commercial use. However, the Duke Breast Cancer MRI dataset is licensed under CC BY-NC 4.0, which restricts commercial use. As this derivative dataset integrates data from all three sources, it must adopt the most restrictive license. Users may freely share and adapt the data with appropriate attribution for non-commercial purposes. For commercial applications involving Duke-derived data, users should contact the original data providers through TCIA.