Statistical shape modeling as a practical tool for dosimetry: quantifying uncertainty in internal dose assessment

To address the recognized limitations of current dosimetry models that rely on average anatomical representations and to enhance the accuracy of internal dose assessments, this study introduces Statistical Shape Modeling (SSM) as a practical and robust tool for applied dosimetry. Our innovative methodology utilizes SSM to construct population-based computational phantoms that accurately reflect the natural anatomical variability within individuals, thereby establishing a more realistic framework for internal dosimetry. Our primary objective is to leverage these newly developed, highly variable phantoms to systematically investigate the uncertainty in Specific Absorbed Fraction (SAF) calculations arising from inter-individual organ size and shape variations. We developed population-based phantoms by integrating organ models derived from SSM into a template phantom. SAF values were subsequently computed using Monte Carlo simulations with the GATE radiation transport code for these phantoms, reflecting variations in lung size. Significant discrepancies in SAF values were observed between the population-based phantoms and the average (template) phantom, highlighting the critical influence of organ size variability. Specifically, at 0.01 MeV photon energy, the percentage differences in SAF values ranged from 19.96% to 54.71%. The promising results of this study demonstrate the efficacy and practicality of employing SSM to generate population-based computational phantoms that more accurately represent anatomical variability. This work uniquely establishes SSM as a viable, practical tool to improve the accuracy of SAF calculations and enhance the reliability of internal dosimetry assessments. Future studies should explore the extension of this methodology to other organs and its application in personalized dosimetry and risk assessment.