Virtual calibration for in vivo measurement of Pb-210 activity in the skull using BOMAB, MIRD, and MIDA phantoms

The counting efficiency calibration for in vivo measurement is crucial to derive the activity of radionuclides residing inside a monitored subject. Recently, virtual calibration based on computational phantoms has become popular, yet some key questions remain unresolved. Here, we focus on the in vivo measurement of Pb-210 in the skull and systematically examine how virtual calibration compares to those using physical phantoms and how the variety of computational phantoms affects the derived counting efficiency. It is found that the virtually calibrated efficiency based on the MIDA phantom, which characterizes the highest anatomical fidelity, shows reasonable consistency with the experimental counterpart, with a relative bias of approximately 10%. However, in comparison to the case based on the MIDA phantom, those based on the BOMAB and MIRD phantoms show larger deviation, demonstrating underestimations on the counting efficiency by 51% and 42%, respectively. This finding underscores the critical role of computational phantoms in the virtual calibration. This study contributes to the development of techniques for assessing lung cancer risk resulting from chronic radon exposure through in vivo measurement of skeletal Pb-210 activity.