The Bi-Gamma ROC Curve in a Straightforward Manner


  • Ehtesham Hussain University of Karachi, Karachi-75270, Pakistan



Sensitivity, Specificity, Receiver Operating Characteristic (ROC) curve, Normal distribution, Gamma distribution


In biomedical research, biomarkers (diagnostic tests) are used in distinguishing healthy and diseased populations. The effectiveness and accuracy of a biomarker generally assessed through the use of a Receiver Operating Characteristic (ROC) curve model, and its functional such as area under the curve (AUC). The parametric (smooth) ROC curves are obtained under the specific distributions assumptions. A resulting ROC curve model is the plot of sensitivity versus 1-specificity for all possible threshold values. Most popular and widely used ROC curve model is bi-normal ROC curve model under the assumptions of normality. When the biomarker results are continuous and positively skewed (non-normal). The gamma distribution is supposed to a flexible model for positively skewed measurements. In practice use of bi-gamma ROC curve model is hindered by the fact that ROC function cannot be written in closed-form. The solution of the problem is to use transformed invariance property of ROC curve model. Which assumes that the test results of both diseased and healthy are normally distributed after some monotone transformation [1]. In this paper we propose a simple approximation solution for the problem mentioned in above lines using a normal approximation due to Wilson and Hilfertys [2]. Which is useful to approximate gamma distribution results with classical normal distribution based results.


Dorfman DD, Alf E. Maximum likelihood estimation of parameters of signal detection theory and determination of confidence intervals-rating-method data. J Math Psych 1969; 6, 487-96.

Wilson EB, Hilferty MM. The Distribution of Chi-Squares, Proc Natl Acad Sci 1931; 17: 684-88.

Metz CE. Some practical issues of experimental design and data analysis in radiological ROC studies. Investig Radiol 1989; 24: 234-45.

Somoza E, Mossman D, McFeeters L. The info-ROC technique: a method for comparing and optimizing inspection systems. In Review of Progress in Quantitative Nondestructive Evaluation, Thomson DO, Chimenti DE (eds). Plenum Press, New York 1990.

Hsiao JK, Barko JJ, Potter WZ. Diagnosing diagnoses: receiver operating characteristic methods and psychiatry. Arch General Psychiatry 1989; 46: 664-67.

Aoki K, Misumi J, Kimura T, Zhao W, Xie T. Evaluation of cutoff levels for screening of gastric cancer using serum pepsinogens and distributions of levels of serum pepsinogens I, II and of PG I/PG II ratios in a gastric cancer case-control study. J Epidemiol 1997; 7: 143-51.

Pepe MS. The Statistical Evaluation of Medical Tests for Classification and Prediction. Oxford Stat Sci Ser 2003.

Begg CB. Advances in statistical methodology for diagnostic medicine in the 1980s. Stat Med 1991; 10: 1887-95.

Zhou H-X, Obuchowski NA, McClish DK. Statistical Methods in Diagnostic Medicine, Wiley, New York 2002.

Pepe MS. The Statistical Evaluation of Medical Tests for Classification and Prediction, Oxford University Press, Oxford 2003.

Bamber DC. The area above the ordinal dominance graph and the area below the receiver operating characteristic graph. J Math Psychol 1975; 12 387-15.

Goddard MJ, Hinberg I. Receiver Operator Characteristic (ROC) Curves and Non-Normal Data: An Empirical Study. Stat Med 1990; 9: 325-37.

Faraggi D, Reiser B. Estimation of the area under the ROC curve. Stat Med 2002; 21: 3093-6.

Faraggi D, Reiser B, Schisterman EF. ROC curve analysis for biomarkers based on pooled assessments. Stat Med 2003; 22: 2515-7.




How to Cite

Ehtesham Hussain. (2012). The Bi-Gamma ROC Curve in a Straightforward Manner. Journal of Basic & Applied Sciences, 8(2), 309–314.