An Online Mammography Database with Biopsy Confirmed Types.

Area of Science:

• Radiology
• Oncology
• Medical Informatics

Background:

• Breast carcinoma is a leading cancer in women globally, with early detection crucial for improving survival rates.
• Mammography is a cost-effective, sensitive imaging tool for early breast cancer diagnosis.
• Existing public mammography datasets lack diversity and comprehensive clinical data, including biopsy confirmation and molecular subtypes.

Purpose of the Study:

• To address the limitations of current datasets by creating a diverse, open-access mammography database.
• To provide a valuable resource for research in breast cancer detection, diagnosis, and treatment.
• To promote advancements in artificial intelligence and machine learning applications in medical imaging.

Main Methods:

• Construction of the Chinese Mammography Database (CMMD) using online breast mammography data.
• The CMMD comprises 3712 mammographies from 1775 patients.
• The database is divided into two subsets: CMMD1 (biopsy-confirmed tumor types) and CMMD2 (known molecular subtypes).

Main Results:

• The CMMD dataset includes 2214 mammographies with biopsy-confirmed benign or malignant tumor types (CMMD1).
• The CMMD dataset includes 1498 mammographies with known molecular subtypes (CMMD2).
• The database enriches the diversity of available mammography data for research.

Conclusions:

• The CMMD represents a significant contribution to the field of breast cancer research by offering a diverse and clinically annotated dataset.
• This database facilitates further development in early breast cancer detection and the study of molecular subtypes.
• The CMMD aims to accelerate research and improve outcomes in breast carcinoma diagnosis and treatment.