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 Video

Updated: Jun 2, 2026

An Experimental Paradigm for the Prediction of Post-Operative Pain (PPOP)
14:56

An Experimental Paradigm for the Prediction of Post-Operative Pain (PPOP)

Published on: January 27, 2010

Machine learning models for predicting postoperative delirium after noncardiac surgery: A comparative study.

Yan Yang1, PengCheng Zhu1

  • 1Department of Anesthesiology, The Second People's Hospital of Hefei, China.

The Journal of International Medical Research
|May 31, 2026
PubMed
Summary

eXtreme Gradient Boosting accurately predicts postoperative delirium risk after noncardiac surgery. This machine learning model offers improved interpretability and performance over traditional methods, aiding in early prevention strategies.

Related Concept Videos

You might also read

Related Articles

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

Sort by
Same journal

Fetomaternal hemorrhage in monochorionic diamniotic twins at 37+ weeks' gestation: A rare complication.

The Journal of international medical research·2026
Same journal

Application of regional anesthesia in pediatric hip surgery for developmental dysplasia of the hip: A retrospective observational study.

The Journal of international medical research·2026
Same journal

The saturation effect of body mass index on femoral neck bone mineral density in a sample of apparently healthy older Syrians: A cross-sectional study.

The Journal of international medical research·2026
Same journal

Development and application of organoid technology in pituitary neuroendocrine tumor research.

The Journal of international medical research·2026
Same journal

Efficacy evaluation of novel targeted drugs in patients with human epidermal growth factor receptor 2-positive breast cancer: A meta-analysis based on multiple sets of clinical data.

The Journal of international medical research·2026
Same journal

S1-4 aptamer-mediated iron oxide nanoparticles for targeted magnetic hyperthermia in MCF-7 breast cancer cells.

The Journal of international medical research·2026

Area of Science:

  • Anesthesiology and Perioperative Medicine
  • Artificial Intelligence in Healthcare
  • Geriatric Medicine and Cognitive Health

Background:

  • Postoperative delirium (POD) is a common, serious complication after noncardiac surgery, associated with adverse outcomes.
  • Existing prediction models face challenges in clinical implementation and interpretability.
  • Machine learning offers potential for improved POD prediction.

Purpose of the Study:

  • To compare the performance of three machine learning models—eXtreme Gradient Boosting (XGBoost), logistic regression, and support vector machine—for early POD prediction.
  • To evaluate the interpretability of these models using Shapley Additive Explanations (SHAP).

Main Methods:

  • Retrospective analysis of 143 adult patients undergoing elective noncardiac surgery.
Keywords:
Machine learningpostoperative delirium

Related Experiment Videos

Last Updated: Jun 2, 2026

An Experimental Paradigm for the Prediction of Post-Operative Pain (PPOP)
14:56

An Experimental Paradigm for the Prediction of Post-Operative Pain (PPOP)

Published on: January 27, 2010

  • Collected 11 perioperative variables, including demographics, clinical scores, and intraoperative parameters.
  • Trained and validated models using an 80:20 data split; assessed performance via AUC, sensitivity, specificity, accuracy, and calibration.
  • Utilized SHAP analysis to identify key predictive features.

    • Main Results:

    • XGBoost demonstrated superior predictive performance (AUC=0.852) compared to logistic regression (0.715) and SVM (0.698).
    • Significant predictors included age, Mini-Mental State Examination score, hemoglobin, surgery duration, opioid dose, lowest mean arterial pressure, blood loss, and ASA class.
    • SHAP analysis highlighted age, MMSE score, and lowest MAP as the most influential features for POD prediction.

    Conclusions:

    • XGBoost provides an accurate and interpretable method for predicting postoperative delirium risk.
    • The model effectively captures complex, nonlinear relationships among predictive variables.
    • This approach can support targeted perioperative interventions to prevent delirium.