MGRegBench: A Novel Benchmark Dataset with Anatomical Landmarks for Mammography Image Registration

2025-12-19

AI Summary

Breast X-ray image registration has long suffered from the absence of public benchmark datasets and standardized evaluation protocols, hindering fair methodological comparisons. To address this, we introduce MGRegBench—the first large-scale, open-source mammographic registration benchmark—comprising over 5,000 paired images, 100 sets of manually annotated anatomical landmark points, and corresponding segmentation masks, enabling robust clinical applications such as disease progression tracking. This is the first benchmark in this modality to provide fine-grained anatomical supervision for unified evaluation. We systematically assess diverse registration paradigms: classical (ANTs), deep learning-based (VoxelMorph, TransMorph), and implicit neural representations (IDIR, MammoRegNet). Evaluation employs both landmark localization error and mask-based metrics (e.g., Dice, Hausdorff distance), yielding the most comprehensive cross-method analysis to date. All data, annotations, and source code are publicly released, substantially enhancing comparability and reproducibility.

Technology Category

Machine Learning: Evaluation and AnalysisComputer Vision: SegmentationData Mining & Knowledge Management: Mining of Visual, Multimedia & Multimodal Data

Application Category

Search and Retrieval-Augmented AI: Web evaluation methodologies and metricsEconomics, Online Markets and Human Computation: Data quality aspects of human-annotated datasetsUser Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

Robust mammography registration is essential for clinical applications like tracking disease progression and monitoring longitudinal changes in breast tissue. However, progress has been limited by the absence of public datasets and standardized benchmarks. Existing studies are often not directly comparable, as they use private data and inconsistent evaluation frameworks. To address this, we present MGRegBench, a public benchmark dataset for mammogram registration. It comprises over 5,000 image pairs, with 100 containing manual anatomical landmarks and segmentation masks for rigorous evaluation. This makes MGRegBench one of the largest public 2D registration datasets with manual annotations. Using this resource, we benchmarked diverse registration methods including classical (ANTs), learning-based (VoxelMorph, TransMorph), implicit neural representation (IDIR), a classic mammography-specific approach, and a recent state-of-the-art deep learning method MammoRegNet. The implementations were adapted to this modality from the authors' implementations or re-implemented from scratch. Our contributions are: (1) the first public dataset of this scale with manual landmarks and masks for mammography registration; (2) the first like-for-like comparison of diverse methods on this modality; and (3) an extensive analysis of deep learning-based registration. We publicly release our code and data to establish a foundational resource for fair comparisons and catalyze future research. The source code and data are at https://github.com/KourtKardash/MGRegBench.

Problem

• Lack of public datasets and benchmarks for mammography image registration.
• Inconsistent evaluation frameworks hinder comparison of registration methods.
• Need for standardized resources to advance mammography registration research.

Innovation

• Created a large public mammography dataset with manual landmarks
• Benchmarked diverse registration methods on mammography images
• Provided open-source code and data for fair research comparisons