MCNP SIMULATIONS WITH A PERSONALISED VOXEL PHANTOM TO VERIFY 131I CONTENT IN THYROID ESTIMATED BASED ON MEASUREMENTS WITH AN NaI(Tl) SPECTROMETER.

One of the authors (O.K.) stayed in the area located ~110 km south from the Fukushima Daiichi Nuclear Power Plant during the arrival of radioactive plumes released into the environment due to the accident in March 2011 in Japan. A previous study determined his 131I thyroid content using an NaI(Tl) spectrometer. The one remaining issue was to investigate the measurement error due to inevitable differences in the configuration (e.g. the thyroid shape and volume) between the physical phantom employed for calibration of the spectrometer and the real subject. In the present study, Monte Carlo simulations for the thyroid measurements were performed using the Monte Carlo N-Particle (MNCP) code to investigate discrepancies in peak efficiencies of the spectrometer between the personalised voxel phantom created from O.K.'s magnetic resonance images and the several typical/reference phantoms that exist. As a result, the peak efficiencies for the Oak Ridge Institute of Nuclear Studies (ORINS) phantom were found to be comparable to those for the reference voxel phantoms reproducing realistic human anatomy (the Adult Male phantom and the Japanese Male phantom). The peak efficiency for the personalised phantom, on the other hand, was fairly close to that of the other physical phantom (the Transfer phantom) actually used for the calibration of the spectrometer, suggesting that the 131I thyroid content determined for him in the previous study was sufficiently accurate. The discrepancies of peak efficiencies at the primal photon energy of 131I (365 keV) among the different phantoms were improved by extending the distance between the neck and the spectrometer; however, the appropriate measurement geometry in an actual situation will depend on the primary purpose of the measurements and should be determined accordingly.