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

Classification algorithms applied to narrative reports.

A Wilcox1, G Hripcsak

  • 1Department of Medical Informatics, Columbia University, New York, NY, USA.

Proceedings. AMIA Symposium
|November 24, 1999
PubMed
Summary
This summary is machine-generated.

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

Epigenetic Signatures in Monozygotic and Dizygotic Twins Discordant for Orofacial Clefts.

medRxiv : the preprint server for health sciences·2026
Same author

Controlled evaLuation of Angiotensin Receptor Blockers for COVID-19 respIraTorY disease (CLARITY): statistical analysis plan for a randomised controlled Bayesian adaptive sample size trial.

Trials·2022
Same author

MAO inhibitory activity of bromo-2-phenylbenzofurans: synthesis, <i>in vitro</i> study, and docking calculations.

MedChemComm·2018
Same author

Estimating summary statistics for electronic health record laboratory data for use in high-throughput phenotyping algorithms.

Journal of biomedical informatics·2018
Same author

A missense mutation in Katnal1 underlies behavioural, neurological and ciliary anomalies.

Molecular psychiatry·2017
Same author

The effects of host plant defoliation and fertilizer application on larval growth and oviposition behaviour in cinnabar moth.

Oecologia·2017
Same journal

Progressive display of very high resolution images using wavelets.

Proceedings. AMIA Symposium·2002
Same journal

The Chronus II temporal database mediator.

Proceedings. AMIA Symposium·2002
Same journal

Gene expression levels in different stages of progression in oral squamous cell carcinoma.

Proceedings. AMIA Symposium·2002
Same journal

An assessment of the visibility of MeSH-indexed medical web catalogs through search engines.

Proceedings. AMIA Symposium·2002
Same journal

Filtering for medical news items using a machine learning approach.

Proceedings. AMIA Symposium·2002
Same journal

Enriching the structure of the UMLS semantic network.

Proceedings. AMIA Symposium·2002
See all related articles

Automated analysis of chest X-ray reports using natural language processing and classification algorithms significantly improves data extraction compared to raw text analysis. Domain knowledge-guided methods outperformed others, highlighting their value in clinical data mining.

Area of Science:

  • Medical informatics
  • Natural Language Processing
  • Machine Learning

Background:

  • Clinical data resides in unstructured narrative reports, limiting automated analysis.
  • Natural language processing (NLP) and data mining offer solutions for extracting this valuable information.
  • Automated decision support systems require structured data, which is often lacking in clinical narratives.

Purpose of the Study:

  • To evaluate multiple classification algorithms for extracting clinical information from chest X-ray reports.
  • To compare the performance of NLP-processed text against raw text for data classification.
  • To identify optimal methods for converting narrative clinical data into a usable format for automated systems.

Main Methods:

  • A general-purpose natural language processor converted narrative chest X-ray reports into coded data.

Related Experiment Videos

  • Six classification methods (rule generation, decision trees, Bayesian classifiers, information retrieval) were applied to 200 reports.
  • Predictor variables were limited to prevent overfitting, with a focus on domain knowledge and conditional probabilities.

    • Main Results:

    • Significant performance differences were observed among the classification algorithms.
    • The best algorithm applied to NLP-processed text outperformed information retrieval on raw text.
    • Methods incorporating domain knowledge demonstrated superior performance compared to those relying solely on conditional probabilities.

    Conclusions:

    • NLP combined with appropriate classification algorithms enhances the extraction of clinical information from narrative reports.
    • Domain knowledge is crucial for developing high-performing predictive models in clinical text analysis.
    • Algorithm performance is influenced by factors such as training set size and variable selection.