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:
|
494
|
Type:
|
Topic Contributed
|
Date/Time:
|
Wednesday, August 1, 2012 : 10:30 AM to 12:20 PM
|
Sponsor:
|
Biopharmaceutical Section
|
Abstract - #304905 |
Title:
|
ANOVA Models for Nontraditional ROC Multi-Reader Imaging Studies
|
Author(s):
|
Stephen Hillis*+
|
Companies:
|
University of Iowa
|
Address:
|
, Iowa City, IA, 52246-2208, United States
|
Keywords:
|
Multi-reader imaging studies ;
ROC ;
correlated ANOVA ;
marginal ANOVA models ;
diagnostic test accuracy
|
Abstract:
|
For diagnostic radiological imaging studies, a common experimental design is the factorial design where each reader read each case under each test. For these studies the outcome of interest is typically a reader-performance measure, such as the area under the receiver-operating characteristic (ROC) curve. Conventional mixed-effects ANOVA procedures cannot be used to account for the correlation between outcomes because these procedures require that the outcome be specific to each test, reader and case combination. The most frequently used statistical method for analyzing data from such a study is the correlated ANOVA method developed by Obuchowski and Rockette (OR) and Dorfman, Berbaum, and Metz (DBM). However, properties of this method have not been derived for other designs. In this talk I demonstrate how OR/DBM can easily be extended and properties derived for other balanced designs by showing that the OR model is equivalent to a marginal ANOVA model corresponding to a conventional ANOVA model. These other designs include split-plot and "mixed" (to be discussed by Nancy Obuchowski) three-factor designs, as well as four-factor designs.
|
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.