JSM 2012 Home

JSM 2012 Online Program

The views expressed here are those of the individual authors and not necessarily those of the JSM sponsors, their officers, or their staff.

Online Program Home

Abstract Details

Activity Number: 511
Type: Contributed
Date/Time: Wednesday, August 1, 2012 : 10:30 AM to 12:20 PM
Sponsor: Biometrics Section
Abstract - #304826
Title: Making Diagnosis Based on Multiple Diagnostic Tests Without a Gold Standard
Author(s): Jingyang Zhang*+ and Ying Zhang and Kathryn Chaloner and Jack T Stapleton
Companies: Fred Hutchinson Cancer Research Center and University of Iowa and University of Iowa and University of Iowa
Address: , Seattle, WA, , United States
Keywords: Diagnostic Testing ; Mixture Model ; ROC analysis

When there is no "gold standard" available, it is common to reconcile information from multiple imperfect diagnostic tests in order to obtain more accuracy. In this paper, we generalize the linear discriminant method and the optimal risk score method to accommodate the situation that is lack of a "gold standard". We also propose an alternative sequential diagnostic method which does not require all tests to be applied to each subject. All the extended methods and the newly proposed method are developed under some parametric distributional assumptions. A mixture of two multivariate normal distributions is used to fit the unclassified data and the optimal diagnostic rule for each method is derived based on the fitted model. We provide the numerical implementation of all methods. Asymptotic results of statistical inferences about the methods are also given. Simulation studies are carried out to compared the methods and the illustration with a real-life data set is included.

The address information is for the authors that have a + after their name.
Authors who are presenting talks have a * after their name.

Back to the full JSM 2012 program

2012 JSM Online Program Home

For information, contact jsm@amstat.org or phone (888) 231-3473.

If you have questions about the Continuing Education program, please contact the Education Department.