Development and Validation of Thalia: A High-Resolution Pediatric Computational Model of a 10-Month-Old Infant

Numerical human models are essential to advance medical device design, safety assessment, and study how anatomical development influences physiological processes. Despite increasing availability of pediatric models, a critical gap remains in high-resolution, non-morphed whole-body models representing children around one year of age. Existing pediatric models are often derived from morphing older anatomies or lack sufficient tissue segmentation to accurately capture early developmental anatomy.

This study introduces Thalia, a non-morphed, high-resolution numerical model of a healthy 10-month-old female. The model was constructed by segmenting 442 tissues from Magnetic Resonance Imaging data. Brain tissues were automatically segmented using an infant-specific FreeSurfer framework, followed by semi-automated and manual refinement in 3DSlicer. The model was validated by expert review and quantitative comparison with age-matched anatomical values reported in the literature. The resulting whole-body model provides detailed anatomical representation across the brain, musculoskeletal system, vasculature, and internal organs, enabling realistic assignment of tissue-specific properties for computational studies. It provides a versatile platform for pediatric medical device development, dosimetry, safety assessment, and bioelectromagnetic simulations. This pediatric numerical anatomical model is openly available as an open-source resource.