Optimization of an in vivo X-ray fluorescence mercury measurement system

A non-invasive in vivo X-ray fluorescence (XRF) method of measuring renal mercury concentrations has previously been reported, as a potential occupational monitoring tool for those who work with this toxic element [Phys. Med. Biol. 40 (1995) 413]. However, the detection limits remain high compared to the typical values anticipated in these populations. Our approach for further enhancing the XRF renal mercury detection limit has been threefold: investigations of the ideal filtration and tube voltage with a conventional tungsten anode X-ray tube, and the replacement of the existing tungsten X-ray tube with a Fluorex tube [Phys. Med. Biol. 36 (1991) 1573]. In all cases the systems were compared by Monte Carlo simulation to that reported by Borjesson et al. [Phys. Med. Biol. 40 (1995) 413]. The optimal filtration was found to be a 0.035 cm uranium filter, positioned after the polarizer. Modest improvement was achieved by increasing the tungsten tube voltage from 160 [Phys. Med. Biol. 40 (1995) 413] to 200 kV, decreasing the system detection limit by 27% for the same subject dose. It was found that the Fluorex tube did not improve the system sensitivity for a given dose rate, either when the tube was used for direct excitation or in a polarized configuration. Despite the improved performance reported here at 200 kV, detection limits remain high compared to typical levels in occupationally exposed individuals. (C) 2003 Elsevier B.V. All rights reserved.