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

Classifying binormal diagnostic tests using separation-asymmetry diagrams with constant-performance curves

E Somoza1

  • 1Psychiatry Service, Department of Veterans Affairs Medical Center, Cincinnati, Ohio 45220.

Medical Decision Making : an International Journal of the Society for Medical Decision Making
|April 1, 1994
PubMed
Summary

This study introduces a novel separation-asymmetry (S-A) diagram for classifying diagnostic tests with binormal distributions. Performance rankings varied significantly between overlap area (iso-OA) and area under the ROC curve (iso-AUR) measures.

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

Cerebrospinal fluid neuroendocrinology of alcohol misusers.

Addiction biology·2016
Same author

What would it take for electrophysiology to become clinically useful?

CNS spectrums·2007
Same author

Metyrapone and cocaine: a double-blind, placebo-controlled drug interaction study.

Pharmacology, biochemistry, and behavior·2005
Same author

Understanding birthing mode decision making using artificial neural networks.

Medical decision making : an international journal of the Society for Medical Decision Making·2002
Same author

The HPA axis in cocaine use: implications for pharmacotherapy.

Journal of addictive diseases·2001
Same author

The measurement of craving.

Journal of addictive diseases·2001

Area of Science:

  • Medical Diagnostics
  • Biostatistics
  • Quantitative Imaging

Background:

  • Diagnostic test performance evaluation is crucial.
  • Binormal distributions commonly model diagnostic test results.
  • Existing classification methods may lack comprehensive performance visualization.

Purpose of the Study:

  • To propose a new method for classifying diagnostic tests based on their underlying binormal distributions.
  • To introduce the separation-asymmetry (S-A) diagram for visualizing test performance.
  • To compare different performance metrics (iso-AUR, iso-OA, iso-MaxInfo) within this framework.

Main Methods:

  • Developed a two-dimensional separation-asymmetry (S-A) diagram using distribution parameters.
  • Superimposed curves of constant performance (iso-AUR, iso-OA, iso-MaxInfo) onto the S-A diagram.

Related Experiment Videos

  • Defined and incorporated the concept of "eccentric" diagnostic tests.

    • Main Results:

    • The S-A diagram allows immediate determination of test performance.
    • Excellent agreement was observed between overlap area (iso-OA) and maximum information (iso-MaxInfo) rankings.
    • Area under the ROC curve (iso-AUR) rankings differed markedly from iso-OA and iso-MaxInfo.
    • Only three of 28 tested diagnostic tests were classified as eccentric.

    Conclusions:

    • The S-A diagram provides a valuable tool for classifying and visualizing diagnostic test performance.
    • Different performance metrics can yield substantially different conclusions about test efficacy.
    • The study highlights the importance of selecting appropriate performance measures for diagnostic tests.