Statistical Evaluation of Diagnostic Performance: Topics in ROC Analysis

摘要

This is an important book in the field of the statistical assessment of the accuracy of diagnostic tests in medicine. Comparable books found in this specialized area of biostatistics include Broemeling (2007), Pepe (2003) and Zhou, Obuchowski & McClish (2002). The book begins with an introduction to measures that estimate test accuracy, including sensitivity, specificity, and area under the receiver operating characteristic (ROC) curve. Also introduced is the Kappa statistic for measuring agreement between a modality and gold standard. The chapter ends with a description of the partial area of the ROC curve and confidence intervals for sensitivity, specificity, and area under the ROC curve.