Abstract
Current dose assessment in nuclear medicine patient studies relies on published S-values, which are, in turn, based on calculated specific absorbed fractions (SAFs) available for a limited number of anthro-pomorphic computational phantoms. In order to take the individual physiognomy of patients more into account, this study aimed to broaden the supply of phantoms and their respective SAFs. An ensemble of 21 mathematical phantoms was submitted to the Monte Carlo Code MCNP4c2 for the purpose of calculation of SAFs for annihilation radiation. These values were incorporated into an internal dose assessment following the Medical Internal Radiation Dose (MIRD) schema and relying on published biokinetic data for intravenous administration of 18F-FDG. The results were compared with data from the ICRP, MIRD reports and concurrent calculations with OLINDA/EXM. A very good agreement with sources relying on the SAFs of Cristy and Eckerman (i.e., the ICRP and OLINDA/EXM) was observed, with the absorbed dose in lung being the only exception. In the case of dose to red marrow, the King Spiers factors were omitted in the three-factor approximation, which led to a precise accordance with the Cristy/Eckerman values. Summarizing, one can say that the coincidence with published data justifies the method chosen and demonstrates successfully the expansion of available reference phantoms for dose assessment in nuclear medicine.