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

Initializing the VA medication reference terminology using UMLS metathesaurus co-occurrences.

John S Carter1, Steven H Brown, Mark S Erlbaum

  • 1University of Utah, USA.

Proceedings. AMIA Symposium
|December 5, 2002
PubMed
Summary

We created a novel algorithm to link medications with diseases they treat using 16 years of data. Over 80% of generated drug-disease pairs were validated by physicians, showing algorithm effectiveness.

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

Suggested Approaches on Communicating to the Public Implications of the Linear No-Threshold (LNT) Dose Response Model and the Fallacy That "There Are No Safe Levels of Radiation Exposure".

Health physics·2026
Same author

Automatic Placement Within a Hierarchical Clinical Decision Support Terminology.

Studies in health technology and informatics·2026
Same author

DeepSeek R1 Distilled Fails to Perform Well Against the USMLE and Other LLMs with and Without Semantics.

Studies in health technology and informatics·2026
Same author

FURIN Stimulates NOTCH2 and NOTCH3 Pathways, Leading to Return of Function in Aged Cells.

Life (Basel, Switzerland)·2026
Same author

Beware the Little Foxes that Spoil the Vines: Small Inconsistencies in Clinical Data Can Distort Machine Learning Findings.

Fortune journal of health sciences·2026
Same author

Semantic Relations: Extending SNOMED CT and Solor.

Applied clinical informatics·2025

Area of Science:

  • Medical Informatics
  • Computational Linguistics
  • Pharmacology

Background:

  • Establishing accurate relationships between medications and diseases is crucial for clinical decision support and pharmacovigilance.
  • Existing methods for identifying drug-disease associations can be labor-intensive and may not capture all potential therapeutic connections.

Purpose of the Study:

  • To develop and evaluate a co-occurrence mining algorithm using the Unified Medical Language System (UMLS) Metathesaurus to identify potential drug-disease treatment pairs.
  • To assess the accuracy and utility of algorithm-generated drug-disease pairs for building a medication reference terminology.

Main Methods:

  • A co-occurrence mining algorithm was developed and applied to 16 years of citation data from the UMLS Metathesaurus.
  • The algorithm generated candidate drug-disease pairs for 100 medication ingredients (50 common, 50 random).

Related Experiment Videos

  • Physician raters evaluated the appropriateness of the generated drug-disease pairs.

    • Main Results:

    • The algorithm produced 977 candidate drug-disease pairs.
    • Over 80% of the candidate pairs were rated as "APPROPRIATE" by physician evaluators.
    • A significant positive correlation was observed between citation frequency and the "APPROPRIATE" rating of drug-disease connections.

    Conclusions:

    • Co-occurrence mining is an effective technique for identifying potential drug-disease treatment relationships.
    • The validated drug-disease pairs are valuable for initializing term definitions in large-scale terminology projects, such as the Veterans Health Administration's medication reference terminology.
    • This approach offers a scalable method for discovering and validating therapeutic associations.