Data governance functions to support responsible data stewardship in pediatric radiology research studies using...

Area of Science:

• Medical Informatics
• Artificial Intelligence in Radiology
• Pediatric Imaging

Background:

• Artificial intelligence (AI) integration is transforming medical practice, particularly in pediatric radiology.
• AI requires large imaging databases for algorithm development and validation.
• Multi-institutional collaboration is essential for robust AI-enabled radiologic research.

Purpose of the Study:

• To define data stewardship and data governance in the context of AI-enabled pediatric radiology research.
• To review key considerations, best practices, and consequences of inadequate data governance.
• To present adaptable data governance frameworks and implementation strategies.

Main Methods:

• Review of data stewardship and data governance principles.
• Analysis of applicability to pediatric radiology research.
• Summarization of data governance frameworks and implementation strategies (centralized vs. distributed).

Main Results:

• Data governance is essential for safeguarding sensitive patient information in large imaging databases.
• Effective governance addresses data privacy, security, bias mitigation, and ethical AI use.
• Both centralized and distributed data management approaches can be implemented using adaptable frameworks.

Conclusions:

• Robust data governance and stewardship are critical for the responsible advancement of AI in pediatric radiology.
• Proper implementation ensures the ethical and secure use of large imaging datasets for AI development.
• Clear frameworks and strategies are needed to facilitate multi-institutional AI-enabled radiologic research.