Answers are grounded in this paper and cite their sources.
Evaluation shows Random Forest and XGBoost effectively estimate radiation in pediatric imaging, suggesting AI's role in ethical decision-making.
The increasing use of imaging techniques in pediatric medicine particularly computed tomography has raised growing concerns about the cumulative risks of ionizing radiation exposure. This study evaluates the predictive capability of machine learning models in estimating the effective radiation dose in pediatric patients, using a synthetic dataset that emulates real-world clinical scenarios. Three supervised regression model Linear Regression, Random Forest, and XGBoost were applied to a dataset of 100 observations, comprising clinical, technical, and ethical variables such as age, imaging modality, and scanned anatomical region, number of repeated exams, informed consent, and medical justification. The Random Forest model demonstrated the best performance (RMSE = 9.89; MAE = 7.78), followed by XGBoost. The most influential variables included age, imaging type, and exam frequency. The findings suggest that artificial intelligence may serve as a valuable tool in optimizing radiological protocols and supporting safer and more ethically sound clinical decision-making. Furthermore, the study underscores the potential of simulation-based methods as ethical and reproducible alternatives for pediatric radiological protection research.
No takes yet. Share an insight, caveat, or question.
Amaya et al. (2025) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: