Improvement of uncertainty by application of MCMC for blood chemical analysis

Abstract

For the improvement of measurement accuracy, Markov Chain Monte Carlo (MCMC) method in BayesianInference has been applied to many fields. We applied this method to type B for practical use on the supplement 1 of ISO-GUM (ISO-guide to express of uncertainty on measurement). Total evaluation is obtained to add the result of the statistical analysis obtained by type A of ISO-GUM to the type B. Combined uncertainty is gotten by carrying out the geometric mean of the results of type A and type B. Assurance value is gotten by multiplying the combined uncertainty by a coverage factor.Key word, MCMC ISO-GIM, uncertainty, PR0001/08/0000-0543 ¥400