Monte Carlo simulation of in vivo measurement of the most suitable knee position for the optimal measurement of activity.

A new computational model has been developed using the Monte Carlo (MC) technique to simulate in vivo measurements with the objective of understanding the most precise measurement location with respect to quantifying the activity of Am in the bones. To benchmark the model, in vivo measurements were performed on the U.S. Transuranium and Uranium Registries (USTUR) case 0846 leg. Front and lateral measurements of the knee of the USTUR case 0846 leg in a bent position and the same measurements with the leg in a straight position using a HP(Ge) detector were completed. Experimental results concluded that the front measurement of the knee in a bent leg position gave the highest count rate, which is an indication of optimal detection efficiency. Therefore, this geometry and knee-detector position were considered as the most appropriate position for knee monitoring. A computational model using MCNPX version 2.6.0 was used to simulate the experimental measurements by using a leg voxel phantom. The mean value and standard deviation (SD) of peak efficiency due to an isotropic 59.5-keV photon from Am were calculated in four different counting geometries. An extra sum of squares F-test was performed on the mean values of the simulated detection efficiencies. The p-value obtained from this statistical test indicates that the differences among the mean values for different counting geometries were significant. These results suggest that the front measurement of a knee in a bent leg position is the optimal counting geometry for in vivo measurement of Am deposited in the bones. The computational model was validated through comparison of the measured and simulated detection efficiencies. It was observed that there is no difference at the 0.1 significant levels between the simulated and measured detection efficiencies in assessment of Am within the bones.