Bias-corrected Hp(10)-to-Organ-Absorbed Dose Conversion Coefficients for the Epidemiological Study of Korean Radiation Workers.

Abstract

The effects of radiation on the health of radiation workers who are constantly susceptible to occupational exposure must be assessed based on an accurate and reliable reconstruction of organ-absorbed doses that can be calculated using personal dosimeter readings measured as and dose conversion coefficients. However, the data used in dose reconstruction contain significant biases arising from a lack of reality and could result in an inaccurate measure of organ-absorbed doses. Therefore, this study quantified the biases involved in organ dose reconstruction and calculated the bias-corrected -to-organ-absorbed dose coefficients for the use in epidemiological studies of Korean radiation workers. Two major biases were considered: (1) the bias in arising from the difference between the dosimeter calibration geometry and the actual exposure geometry and (2) the bias in air kerma-to- conversion coefficients resulting from geometric differences between the human body and slab phantom. The biases were quantified by implementing personal dosimeters on the slab and human phantoms coupled with Monte Carlo method and considered to calculate the bias-corrected -to-organ-absorbed dose conversion coefficients. The bias in was significant for large incident angles and low energies (e.g., 0.32 for right lateral at 218 keV), whereas the bias in dose coefficients was significant for the posterior-anterior (PA) geometry only (e.g., 0.79 at 218 keV). The bias-corrected -to-organ-absorbed dose conversion coefficients derived in this study were up to 3.09-fold greater than those from ICRP publications without considering the biases. The obtained results will aid future studies in assessing the health effects of occupational exposure of Korean radiation workers. The bias-corrected dose coefficients of this study can be used to calculate organ doses for Korean radiation workers based personal dose records.