Mesh-based detailed skeletal models for the ICRP reference pediatric individuals: development and dosimetric implications.

Abstract

<i>Objective.</i>The skeleton contains the red bone marrow (RBM) and the endosteum, tissues linked to radiation-induced leukemia and bone cancer, making their consideration essential in radiation dosimetry. Although adult skeletal dosimetry has advanced with 3D images such as<i>μ</i>CT images, the scarcity of comparable pediatric images prevents pediatric skeletal dosimetry from achieving a similar level. This study aims to develop 3D image-based detailed pediatric skeletal models that, while grounded in adult<i>μ</i>CT images, incorporate the anatomical features of the developing pediatric skeleton.<i>Approach.</i>Target skeletal values were established from extensive anatomical literature and International Commission on Radiological Protection publications, including skeletal tissue mass, cellularity factor, trabecular bone volume fraction, and trabecular number. Guided by these values, trabecular bone models converted from adult<i>μ</i>CT images were refined, a 50<i>μ</i>m endosteal layer was defined, yellow bone marrow (YBM) was incorporated as adipocytes, and remaining regions were assigned as RBM. All modeling steps were performed automatically using our C++-based bone modeling program.<i>Main results.</i>A total of 246 pediatric skeletal models were developed in a high-quality mesh format across six age and sex groups (sex-averaged newborn, 1 year-old, 5 year-old, and 10 year-old, and sex-specific 15 year-old male and female), with each group comprising 41 models. These models represent trabecular bone and RBM/YBM in both the shallow and deep marrow, and all matched their target values within 2%. For selected cases, PHITS Monte Carlo simulations were used to calculate specific absorbed fractions, which increased with decreasing age due to differences in target mass and the combined effects of the anatomical factors incorporated in this study.<i>Significance.</i>This study provides the first comprehensive set of 3D image-based pediatric skeletal models for skeletal dosimetry. These models, together with the dosimetric datasets derived from them, are expected to provide an anatomically robust foundation for improving the accuracy and reliability of pediatric skeletal dosimetry.