Nonparametric ROC summary statistics for correlated diagnostic marker data

We propose efficient nonparametric statistics to compare medical imaging modalities in multi-reader multi-test data and to compare markers in longitudinal ROC data. The proposed methods are based on the weighted area under the ROC curve, which includes the area under the curve and the partial area under the curve as special cases. The methods maximize the local power for detecting the difference between imaging modalities. We develop the asymptotic results of the proposed methods under a complex correlation structure. Our simulation studies show that the proposed statistics result in much better powers than existing statistics. We apply the proposed statistics to an endometriosis diagnosis study.

[1]  Edgar Brunner,et al.  Nonparametric analysis of clustered data in diagnostic trials : Estimation problems in small sample sizes , 2008 .

[2]  Carola Werner,et al.  Rank methods for the analysis of clustered data in diagnostic trials , 2007, Comput. Stat. Data Anal..

[3]  Joseph Stanford,et al.  Incidence of endometriosis by study population and diagnostic method: the ENDO study. , 2011, Fertility and sterility.

[4]  Jialiang Li,et al.  Weighted area under the receiver operating characteristic curve and its application to gene selection , 2010, Journal of the Royal Statistical Society. Series C, Applied statistics.

[5]  R. Serfling.  Approximation Theorems of Mathematical Statistics , 1980 .

[6]  Gary Longton,et al.  Incorporating the Time Dimension in Receiver Operating Characteristic Curves: A Case Study of Prostate Cancer , 1999, Medical decision making : an international journal of the Society for Medical Decision Making.

[7]  Mei-Ling Ting Lee,et al.  The average area under correlated receiver operating characteristic curves : a nonparametric approach based on generalized two-sample Wilcoxon statistics , 2001 .

[8]  Mitchell H. Gail,et al.  A family of nonparametric statistics for comparing diagnostic markers with paired or unpaired data , 1989 .

[9]  Gang Li,et al.  A Unified Approach to Nonparametric Comparison of Receiver Operating Characteristic Curves for Longitudinal and Clustered Data , 2008, Journal of the American Statistical Association.

[10]  D. Bamber.  The area above the ordinal dominance graph and the area below the receiver operating characteristic graph , 1975 .

[11]  Lee-Jen Wei,et al.  Combining dependent tests with incomplete repeated measurements , 1985 .

[12]  K. Zou,et al.  Comparison of correlated receiver operating characteristic curves derived from repeated diagnostic test data. , 2001, Academic radiology.

[13]  Stuart G. Baker,et al.  A Proposed Design and Analysis for Comparing Digital and Analog Mammography , 2001 .

[14]  B. Reiser,et al.  Comparing the Areas Under Two Correlated ROC Curves: Parametric and Non‐Parametric Approaches , 2006, Biometrical journal. Biometrische Zeitschrift.

[15]  M. Canis,et al.  Revised American Society for Reproductive Medicine classification of endometriosis: 1996. , 1997, Fertility and sterility.

[16]  Jean L Freeman,et al.  A non-parametric method for the comparison of partial areas under ROC curves and its application to large health care data sets. , 2002, Statistics in medicine.

[17]  Xiao-Hua Zhou,et al.  Statistical Methods in Diagnostic Medicine , 2002 .

[18]  N. Obuchowski,et al.  Hypothesis testing of diagnostic accuracy for multiple readers and multiple tests: An anova approach with dependent observations , 1995 .

[19]  Yuqing Yang,et al.  Combining dependent tests to compare the diagnostic accuracies--a non-parametric approach. , 2006, Statistics in medicine.

[20]  J. Copas,et al.  Overestimation of the receiver operating characteristic curve for logistic regression , 2002 .

[21]  N A Obuchowski,et al.  Nonparametric analysis of clustered ROC curve data. , 1997, Biometrics.

[22]  E. DeLong,et al.  Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach. , 1988, Biometrics.

[23]  Xiao Song,et al.  A marginal model approach for analysis of multi-reader multi-test receiver operating characteristic (ROC) data. , 2005, Biostatistics.