Artificial Intelligence for Radiation Dose Optimization in Pediatric Radiology: A Systematic Review

Area of Science:

• Medical Imaging
• Radiology
• Artificial Intelligence

Background:

• Pediatric radiology requires careful radiation dose optimization due to children's increased sensitivity to ionizing radiation.
• Existing research on artificial intelligence (AI) for dose optimization in pediatric computed tomography (CT) is limited, with only one prior narrative review.

Purpose of the Study:

• To systematically review AI techniques and architectures used for radiation dose optimization in pediatric radiology.
• To identify specific application areas and evaluate the performance of these AI methods.

Main Methods:

• A systematic literature search was conducted on electronic databases on June 3, 2022.
• Sixteen articles meeting predefined selection criteria were included in the review.

Main Results:

• Deep convolutional neural networks (CNNs) were the predominant AI technique and architecture employed.
• AI demonstrated the potential to reduce radiation doses by 36-70% across various CT examinations (abdomen, chest, head, neck, pelvis) without compromising diagnostic information.
• Both commercial and homegrown AI models showed promising and comparable performance in dose optimization.

Conclusions:

• AI, particularly deep CNNs, shows significant promise for radiation dose optimization in pediatric radiology.
• Further research is needed to explore AI's value across a broader range of modalities and examination types due to the limited scope of current studies.