Role of Bayesian Inference in TLD-Based Personnel Monitoring

In the realm of personnel dosimetry using TLD badges, the analysis of data for epidemiological studies involving radiation-exposed workers has traditionally leaned heavily on recorded annual radiation doses. However, despite the assumption of precise dose values, it's widely recognized that these datasets inherently contain uncertainties stemming from factors such as measurement errors, dose estimation algorithms, censoring, and dose rounding. In this chapter, set within the context of TLD personnel monitoring, a probabilistic framework is applied to elucidate an individual's monthly radiation dose. This endeavour involved the development of statistical methodologies aimed at estimating the actual doses using the reported occupational dose data from TLD personnel monitoring laboratories. Importantly, the chapter unveils a consistent pattern of underestimation of doses for workers, particularly within the lower dose range. This revelation highlights the challenges in attaining accurate estimations of dose–response relationships and the associated standard errors. Consequently, it prompts a critical examination of the adequacy of relying solely on annually recorded doses when conducting studies involving radiation-exposed workers.