[Analysis and Enlightenment of Foreign Open Breast Image Datasets]

Area of Science:

• Medical Artificial Intelligence
• Computer Vision in Healthcare
• Medical Imaging Analysis

Background:

• The increasing prevalence of medical artificial intelligence (AI) has highlighted the potential of computer vision for analyzing medical images.
• Computer vision aids clinicians in disease identification and screening, making it a key research area.
• Access to diverse datasets is crucial for developing robust AI models in medicine.

Purpose of the Study:

• To review and summarize the characteristics, specific details, and related research of international breast image datasets.
• To explore methods for data sharing and utilization of these foreign datasets.
• To provide insights and inspiration for the development and opening of Chinese medical and health data resources.

Main Methods:

• Systematic literature review of foreign breast image datasets.
• Analysis of dataset characteristics, including size, acquisition methods, and annotation quality.
• Examination of existing research utilizing these datasets.
• Investigation of data sharing policies and access mechanisms.

Main Results:

• Detailed overview of several prominent international breast image datasets.
• Identification of common challenges and best practices in dataset curation and sharing.
• Examples of successful applications of computer vision in breast cancer screening using these datasets.
• Insights into the potential for data standardization and collaborative research.

Conclusions:

• International breast image datasets offer valuable resources for advancing medical AI and computer vision applications.
• Effective data sharing and utilization strategies are essential for accelerating research and clinical translation.
• This review provides a foundation for developing similar initiatives within China's medical data landscape.