MAMA-MIA: A Large-Scale Multi-Center Breast Cancer DCE-MRI Benchmark Dataset with Expert Segmentations

2024-06-19

arXiv.org

AI Summary

AI research on breast cancer DCE-MRI is hindered by the scarcity of high-quality expert annotations. To address this, we introduce the largest publicly available multicenter DCE-MRI benchmark dataset for breast cancer to date, comprising 1,506 cases. For the first time, all cases feature expert-level corrected annotations covering both primary tumors and non-mass enhancement regions, alongside 49 structured clinical variables and pre-trained nnU-Net weights. We propose a semi-automatic annotation pipeline integrating nnU-Net–based initial segmentation, radiologist verification, and multi-expert consensus, combined with rigorous image standardization and cross-center harmonization to ensure scalability and reliability. The dataset enables robust model development and fair benchmarking. Baseline nnU-Net achieves Dice scores of 0.82 (tumor) and 0.74 (non-mass) on the held-out test set, demonstrating strong practical utility and generalizability across centers and lesion types.

Technology Category

• Computer Vision: Segmentation
• Machine Learning: Calibration & Uncertainty Quantification
• Knowledge Representation and Reasoning: Diagnosis and Abductive Reasoning

Application Category

• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Responsible Web: Data and user privacy-enhancing technologies for the Web

Abstract

Artificial Intelligence (AI) research in breast cancer Magnetic Resonance Imaging (MRI) faces challenges due to limited expert-labeled segmentations. To address this, we present a multicenter dataset of 1506 pre-treatment T1-weighted dynamic contrast-enhanced MRI cases, including expert annotations of primary tumors and non-mass-enhanced regions. The dataset integrates imaging data from four collections in The Cancer Imaging Archive (TCIA), where only 163 cases with expert segmentations were initially available. To facilitate the annotation process, a deep learning model was trained to produce preliminary segmentations for the remaining cases. These were subsequently corrected and verified by 16 breast cancer experts (averaging 9 years of experience), creating a fully annotated dataset. Additionally, the dataset includes 49 harmonized clinical and demographic variables, as well as pre-trained weights for a baseline nnU-Net model trained on the annotated data. This resource addresses a critical gap in publicly available breast cancer datasets, enabling the development, validation, and benchmarking of advanced deep learning models, thus driving progress in breast cancer diagnostics, treatment response prediction, and personalized care.

Problem

• Limited expert-labeled breast cancer MRI segmentations
• Development of a multicenter breast cancer MRI dataset
• Enabling advanced AI model benchmarking

Innovation

• Deep learning for preliminary segmentation
• Expert-verified multicenter MRI dataset
• Pre-trained nnU-Net model included