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

Development and Validation of a Machine Learning Tool for Plastic Surgery Residency Application Screening.

Katherine J Zhu1, Preetham Bachina1, Matthew J Heron1

  • 1Department of Plastic and Reconstructive Surgery, Johns Hopkins School of Medicine, Baltimore, Maryland.

Journal of Surgical Education
|February 4, 2026
PubMed
Summary
This summary is machine-generated.

Related Concept Videos

A machine learning (ML) tool can efficiently screen plastic surgery residency applications. This AI-powered system helps identify candidates for interviews, improving the application review process.

Area of Science:

  • Medical education
  • Artificial intelligence in medicine
  • Surgical residency selection

Background:

  • Integrated plastic surgery residency programs face increasing application numbers.
  • Faculty review workload is growing due to more applications per position.
  • Artificial intelligence (AI) offers a solution for efficient and holistic application review.

Purpose of the Study:

  • To develop and validate a machine learning (ML) tool for screening residency applications.
  • To identify candidates likely to receive interview invitations using ML.

Main Methods:

  • Retrospective collection of applications from an integrated plastic surgery residency program (2022-2025).
  • Processing application data through four ML models: XGBoost, Random Forest, CatBoost, and LightGBM.
Keywords:
artificial intelligencemachine learningmedical educationresidency admissions

Related Experiment Videos

  • Training and validation on 2022-2024 data, with testing on 2025 data; performance assessed against faculty interview decisions.

    • Main Results:

    • The CatBoost algorithm demonstrated top performance with an AUROC of 0.92 and AUPRC of 0.668.
    • Sensitivity reached 95% and specificity 67% at an F1 score-maximizing threshold.
    • The total number of publications was the most significant feature influencing interview invitation decisions.

    Conclusions:

    • A validated ML algorithm can accurately assist in selecting residency interviewees.
    • This AI tool can enhance the efficiency and holism of residency application reviews.