Divide and Conquer: A Large-Scale Dataset and Model for Left-Right Breast MRI Segmentation

2025-07-18

AI Summary

A publicly available, large-scale breast MRI dataset with precise left/right breast segmentation annotations is currently lacking, and no deep learning model has been specifically designed for this task—severely hindering breast symmetry analysis and clinical decision support. Method: We introduce BraMIS-13K, the first open-source, large-scale breast MRI dataset for left/right segmentation, comprising 13,248 multi-center scans. We further propose SAM-Net, a lightweight symmetry-aware segmentation network incorporating standardized preprocessing, spatial left/right prior modeling, and rigorous multi-center cross-validation. Contribution/Results: On an independent test set, SAM-Net achieves Dice scores of 0.962 (left breast) and 0.965 (right breast). Both the BraMIS-13K dataset and pre-trained SAM-Net models are fully open-sourced. This work significantly advances automated breast MRI segmentation accuracy and cross-site generalizability, establishing a foundational infrastructure for computational breast imaging research and women’s health analytics.

Technology Category

• Computer Vision: Segmentation
• Machine Learning: Multi-instance/Multi-view Learning
• Humans and AI: Brain-Sensing and Analysis

Application Category

• Web Mining and Content Analysis: Large pretrained models with web data
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets

Abstract

We introduce the first publicly available breast MRI dataset with explicit left and right breast segmentation labels, encompassing more than 13,000 annotated cases. Alongside this dataset, we provide a robust deep-learning model trained for left-right breast segmentation. This work addresses a critical gap in breast MRI analysis and offers a valuable resource for the development of advanced tools in women's health. The dataset and trained model are publicly available at: www.github.com/MIC-DKFZ/BreastDivider

Problem

• Lack of public breast MRI dataset with left-right labels
• Need for accurate left-right breast segmentation model
• Limited resources for advanced women's health tools

Innovation

• First public breast MRI dataset with left-right labels
• Deep-learning model for left-right breast segmentation
• Over 13,000 annotated cases for training