THE APPLICATION OF BAYESIAN TECHNIQUES IN THE INTERPRETATION OF BIOASSAY DATA

摘要

The inverse problem of internal dosimetry is naturally posed as a problem of Bayesian inference. The Bayesian approach is of practical importance in three areas: (1) avoiding false positives in the detection of rare events, (2) the calculation of uncertainties, and (3) the calculation of multiple intakes, all of which are important for internal dosimetry. In this paper, the Bayesian approach to the interpretation of measurements is first reviewed using a simple conceptual example. Then, a simple ~(239)Pu case using IMBA expert is discussed, and finally a current cutting-edge example is discussed involving real ~(238)Pu data calculated with a Markov Chain Monte Carlo algorithm and with exact calculation of poisson likelihood functions.