Main description:

Offering a detailed and careful explanation of the methods, this book delineates Bayesian non parametric techniques to be used in health care and the statistical evaluation of diagnostic tests to determine accuracy before mass use in practice. Unique to these methods is the incorporation of prior information and elimination of subjective beliefs and asymptotic results. It includes examples such as ROC curves and ROC surfaces estimation, modeling of multivariate diagnostic data, absence of a perfect test, ROC regression methodology, and sample size determination.

Contents:

Introduction. An overview of diagnostic tests concepts. Bayeisan nonparametric (BNP) methods. ROC curve estimation. More on ROC estimation (stochastic ordering, modeling multivariate data, ROC surfaces). BNP regression models. Absence of gold standard. Sample size determination and prevalence estimation. Appendix: Getting started with WinBUGS and R.