JoVE - Journal of Visualized Experiments
JoVE Visualize
Contact Us

Machine learning model predicts airway stenosis requiring clinical intervention in patients after lung transplantation: a retrospective case-controlled study

  • 0Department of Thoracic Surgery, West China Hospital, Sichuan University, Chengdu, 610041, China. 22tiandong@163.com.
Bmc Medical Informatics and Decision Making +

|

August 19, 2024

|

August 19, 2024

View abstract on PubMed

Summary

This summary is machine-generated.

Machine learning models effectively predict airway stenosis (AS) after lung transplantation (LTx), outperforming traditional methods. This advancement aids in early intervention for lung transplant recipients.

Area Of Science

  • Medical Informatics
  • Pulmonary Medicine
  • Machine Learning Applications

Background

  • Airway stenosis (AS) poses significant risks, including morbidity and mortality, for patients following lung transplantation (LTx).
  • Accurate prediction of AS is crucial for timely clinical intervention in post-LTx patients.

Purpose Of The Study

  • To develop and validate machine learning (ML) models for predicting airway stenosis (AS) in lung transplantation (LTx) recipients.
  • To compare the predictive performance of ML models against conventional logistic regression (LR).

Main Methods

  • Retrospective review of 381 LTx patients (2017-2019).
  • Development of conventional logistic regression (LR) and 56 ML models using various feature selection and algorithms.
  • Internal validation using bootstrap method; performance assessed by Area Under the Curve (AUC) and Brier score.

Main Results

  • 40 (10.5%) patients developed AS. Male sex, pulmonary arterial hypertension, and postoperative 6-minute walking test were significant predictors.
  • The optimal ML model (Random Forest algorithm) achieved an AUC of 0.760 and Brier score of 0.085.
  • The optimal ML model demonstrated superior performance compared to the conventional LR model (AUC 0.689, Brier score 0.091).

Conclusions

  • An optimal machine learning model, utilizing clinical characteristics, can satisfactorily predict airway stenosis in lung transplant patients.
  • This ML approach offers a promising tool for improving patient outcomes after lung transplantation.

Keywords: