Use of Artificial Intelligence in Radiology: Impact on Pediatric Patients, a White Paper From the ACR Pediatric AI...

Area of Science:

• Radiology
• Medical Informatics
• Artificial Intelligence in Healthcare

Background:

• The widespread adoption of artificial intelligence (AI) in medical imaging presents unique challenges for pediatric applications.
• A significant health equity issue exists due to the underrepresentation and inadequacy of AI development for children.
• Current AI development and validation methods, often designed for adults, may not be suitable or safe for pediatric populations.

Purpose of the Study:

• To educate the radiology community on the critical need for pediatric-specific artificial intelligence (AI).
• To enhance understanding of the distinct issues surrounding AI design, training, validation, and implementation in children.
• To propose actionable solutions to overcome the current inadequacies in pediatric AI development and ensure equitable access to AI-driven healthcare.

Main Methods:

• A white paper authored by the ACR Pediatric AI Workgroup of the Commission on Informatics.
• Review and synthesis of existing knowledge on AI in medical imaging, with a focus on pediatric considerations.
• Development of recommendations for the specialized approaches required for pediatric AI.

Main Results:

• Identified a critical gap in pediatric AI, highlighting a significant health equity concern.
• Emphasized that the design, training, validation, and implementation of AI for children necessitate distinct methodologies compared to adults.
• Provided a framework and call to action for the radiology community to address these inadequacies.

Conclusions:

• The development and deployment of AI in pediatric radiology demand specific, careful, and distinct approaches.
• Addressing the lack of pediatric AI is crucial for ensuring health equity in AI-driven medical imaging.
• The radiology community is urged to collaborate and adopt specialized strategies to guarantee safe, reliable, and effective AI for children, referencing the Image IntelliGently initiative.