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

Evaluating variable selection methods for diagnosis of myocardial infarction.

S Dreiseitl1, L Ohno-Machado, S Vinterbo

  • 1Harvard Medical School/Massachusetts Institute of Technology, Division of Health Sciences and Technology, Boston, USA. sdreisei@dsg.harvard.edu

Proceedings. AMIA Symposium
|November 24, 1999
PubMed
Summary

This study compared machine learning methods for predicting myocardial infarction. While some variables were consistently identified as important, no single predictor was selected by all models, highlighting variability in variable selection techniques.

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

Training in Health Informatics in Brazil.

Yearbook of medical informatics·2016
Same author

Creating a Common Data Model for Comparative Effectiveness with the Observational Medical Outcomes Partnership.

Applied clinical informatics·2015
Same author

Health IT and clinical decision support systems: human factors and successful adoption.

Journal of the American Medical Informatics Association : JAMIA·2014
Same author

"Big data" and the electronic health record.

Yearbook of medical informatics·2014
Same author

Confluence of disciplines in health informatics: an international perspective.

Methods of information in medicine·2011
Same author

Feasibility evaluation of Smart Stretcher to improve patient safety during transfers.

Methods of information in medicine·2010

Area of Science:

  • Cardiology
  • Machine Learning
  • Data Science

Background:

  • Predicting myocardial infarction (MI) is crucial for timely intervention.
  • Identifying key predictive variables aids in developing effective diagnostic and prognostic tools.
  • Machine learning offers powerful methods for analyzing complex health datasets.

Purpose of the Study:

  • To evaluate the performance of various machine learning algorithms in variable selection for myocardial infarction prediction.
  • To identify consensus and discrepancies in predictor variable relevance across different ML techniques.
  • To determine the most consistently relevant input variables for MI prediction from a set of 43.

Main Methods:

  • Investigated logistic regression (stepwise, forward, backward selection), backpropagation for multilayer perceptrons, Bayesian neural networks, and rough sets for variable selection.

Related Experiment Videos

  • Utilized self-organizing maps as an independent method for evaluating and visualizing selected predictor variable subsets.
  • Applied these methods to a myocardial infarction dataset with 43 input variables.

    • Main Results:

    • Multiple machine learning algorithms were applied to identify relevant predictors for myocardial infarction.
    • Good agreement was observed for some selected variables across different methods.
    • Significant variability existed in variable selection outcomes, with only one shared predictor identified across all models.

    Conclusions:

    • Machine learning techniques show promise in identifying relevant variables for myocardial infarction prediction.
    • The choice of algorithm influences the selection of predictive variables, indicating a need for careful method selection.
    • Further research is warranted to consolidate findings and establish a robust set of universally agreed-upon predictors for MI.