Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Multireader, multicase receiver operating characteristic methodology: a bootstrap analysis

D D Dofrman1, K S Berbaum, R V Lenth

  • 1Department of Psychology, University of Iowa, Iowa City 52242, USA.

Academic Radiology
|July 1, 1995
PubMed
Summary

Bootstrapping methods, including case-resampling and reader-case resampling, validate the Dorfman-Berbaum-Metz (DBM) and Toledano-Gatsonis (TG) receiver operating characteristic (ROC) methods. These approaches offer insights into precision and power trade-offs in reader studies.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Multiple diagnostic task performance in CT examination of the chest.

The British journal of radiology·2012
Same author

Multiple diagnostic task performance in CT examination of the chest.

The British journal of radiology·2012
Same author

Statistical power calculations.

Journal of animal science·2006
Same author

Satisfaction of search in osteoradiology.

AJR. American journal of roentgenology·2001
Same author

Factors influencing visualization of vertebral metastases on MR imaging versus bone scintigraphy.

AJR. American journal of roentgenology·2001
Same author

Gaze dwell times on acute trauma injuries missed because of satisfaction of search.

Academic radiology·2001

Area of Science:

  • Medical Imaging Analysis
  • Statistical Methods in Research

Background:

  • Receiver operating characteristic (ROC) analysis is crucial for evaluating diagnostic test performance.
  • The Dorfman-Berbaum-Metz (DBM) and Toledano-Gatsonis (TG) methods are established statistical techniques for analyzing ROC data.
  • Bootstrapping offers a robust approach for validating statistical conclusions.

Purpose of the Study:

  • To evaluate the validity of the DBM and TG methods using bootstrapping techniques.
  • To compare conclusions derived from bootstrapping with those from DBM and TG methods.
  • To investigate the impact of different bootstrapping strategies on the reliability of ROC analysis.

Main Methods:

  • Two bootstrapping methods were employed: resampling cases with fixed readers, and resampling both cases and readers.

Related Experiment Videos

  • The study utilized a well-known dataset for comparative analysis.
  • Analysis of variance (ANOVA) of pseudovalues was used to assess reader variance and treatment interactions.

    • Main Results:

    • Case-resampling bootstrap yielded highly similar results to the DBM and TG methods for individual and average reader performance.
    • When reader variance and random interactions were negligible, bootstrap and DBM/TG methods converged.
    • Both bootstrapping methods indicated smaller standard errors for group means compared to individual reader means, suggesting a precision-power trade-off.

    Conclusions:

    • Case-resampling bootstrap provides statistical justification for the DBM and TG methods in ROC analysis.
    • The findings support the reliability of DBM and TG methods when reader variance is minimal.
    • Evidence suggests a trade-off between readers and cases impacting precision and power in the analyzed dataset.