Catalog of available software

Click the checkbox(es) of the Operating System(s) you want to use for the program(s) you have chosen.
Then verify your choice by selecting "cart" or "checkout" to receive your software.
Program Brief Description Operating System
ROCKIT

[0.9.1 Beta]

 Fits a conventional binormal ROC curve to continuously-distributed or categorical data. Also calculates the statistical significance of differences between ROC estimates obtained from fully paired, partially paired, or unpaired datasets that are associated with 2 "treatments" (e.g, imaging or image-reading conditions). The statistical significance tests performed by this program take case-sample variation but not reader variation into account, so inferences drawn from them can be generalized to a population of cases but not to a population of readers. After the registration process a new version (1.1 Build 2) can also be downloaded (Windows and OS X native). A completely new build of ROCKIT is in preparation and should be released around the end of 2008
DBM MRMC

[2.2 Build 3]

 New software that essentially replaced LABMRMC. Uses the Dorfman-Berbaum-Metz algorithm to compare multiple readers and multiple treatments (e.g., "imaging modalities"). This program employs jackknifing and ANOVA techniques to test the statistical significance of differences between treatments and between readers. Build 3 corrects some issues with partial areas and jackknifing (June 24th 2008) NOT YET AVAILABLE
LABMRMC

[1.0.3 Beta]

 Uses the Dorfman-Berbaum-Metz algorithm to compare multiple readers and multiple treatments (e.g., "imaging modalities"). This program employs jackknifing and ANOVA techniques to test the statistical significance of differences between treatments and between readers.
PLOTROC.xls

[1.0.0 ]

 This is a Microsoft Excel 5.0 spreadsheet macro that takes values of the a and b parameters of the conventional binormal model as input and plots an ROC curve suitable for presentation and publication.
ROCPWR

[6.93 Beta]

 Computes the statistical power of the statistical tests performed by ROCKIT given estimates of the a parameters, b parameters, and correlation parameters of the 2 ROC curves in question. Appropriate values of these parameters can be obtained by guesswork or by using ROCKIT to analyze pilot-study data. ROCPWR outputs a table of the estimated statistical power for a variety of "alpha levels" and case sample sizes. NOT AVAILABLE