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

Feature selection methodology based on explainable AI: application in predicting extubation success in the intensive

JaeBin Sung1, Geun-Hyeong Kim1, Jae-Woo Kim1

  • 1Medical Artificial Intelligence Center, Chungbuk National University Hospital, Cheongju, Republic of Korea.

Scientific Reports
|June 11, 2026
PubMed
Summary

Related Concept Videos

Endotracheal Tube Extubation01:24

Endotracheal Tube Extubation

Endotracheal tube extubation is a critical procedure in weaning patients from mechanical ventilation. It involves physically removing the oral or nasal endotracheal (ET) tube, marking the final step in liberating a patient from ventilatory support.
Procedure
Extubation removes the endotracheal tube (ETT) from the patient on mechanical ventilation. It requires a well-coordinated, multidisciplinary approach involving physicians, nurses, respiratory therapists, and other healthcare professionals.

You might also read

Related Articles

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

Sort by
Same author

Blood Eosinophil Count, Disease Severity, and Exacerbation Risk in Korean Patients with Bronchiectasis: A Retrospective Cohort Study.

Tuberculosis and respiratory diseases·2026
Same author

Clinical variable-based decision-support model for rapid differentiation of hemorrhagic and ischemic stroke at emergency department presentation in South Korea.

Acute and critical care·2026
Same author

Lacticaseibacillus parahuelsenbergensis THGS-36 alleviates acne-like inflammation and lipogenesis in sebocytes via CREB signaling.

Life sciences·2026
Same author

Leuconostoc mesenteroides TH-ST373 Cell-Free Supernatant Against Multidrug-Resistant Escherichia coli.

Probiotics and antimicrobial proteins·2026
Same author

Racial Differences in Patterns of Long-Term Conditions Among Adults With Asthma.

Clinical and experimental allergy : journal of the British Society for Allergy and Clinical Immunology·2026
Same author

Dose-dependent biphasic effect of environmental UVA on stem cell function through PRPF40A, TGF-β1, NFATc1 signaling.

Journal of photochemistry and photobiology. B, Biology·2026

This study introduces a new explainable AI (XAI) feature selection (FS) method for medical data. It enhances predictive model interpretability and performance, particularly in intensive care unit (ICU) settings.

Area of Science:

  • Artificial Intelligence
  • Medical Informatics
  • Machine Learning

Background:

  • High-dimensional medical data present challenges for predictive modeling due to noise and the curse of dimensionality.
  • Robust feature selection (FS) is crucial for developing reliable predictive models in healthcare.

Purpose of the Study:

  • To develop and evaluate a novel, model-agnostic feature selection (FS) method using explainable AI (XAI) techniques.
  • To improve the interpretability and robustness of predictive models for high-dimensional medical data.

Main Methods:

  • Proposed a consensus feature ranking by aggregating SHAP, LIME, and LRP explainers.
  • Reordered features using true-positive (TP)-specific attributions.
  • Applied the method to predict extubation success in 19,567 mechanically ventilated patients using the MIMIC-IV database and a multilayer perceptron.
Keywords:
E0xtubation predictionExplainable artificial intelligence (XAI)Feature selectionHigh-dimensional dataModel interpretability

Related Experiment Videos

Main Results:

  • The XAI FS method achieved an AUROC of 0.8527 and comparable sensitivity to conventional methods.
  • The selected feature subset was compact, interpretable, and clinically plausible, aligning with established extubation readiness criteria.
  • The framework demonstrated potential generalizability, improving performance on most classification tasks across seven additional medical datasets.

Conclusions:

  • True-positive focused, multi-explainer FS provides robust and interpretable models for high-risk ICU decision support.
  • This approach establishes reliable consensus feature rankings, enhancing clinical utility.
  • The method shows promise for broader application in medical predictive modeling.