Related Experiment Videos

Machine learning-driven risk assessment of severe Mycoplasma pneumoniae in children: analysis based on core clinical

Duoduo Li1, Li Wang1, Xiaolu Zhao2

  • 1Department of Pediatrics, The First Affiliated Hospital of Henan Medical University, Weihui, Henan, China.

Frontiers in Medicine
|August 5, 2026
PubMed

Insights

A machine learning model accurately predicts severe Mycoplasma pneumoniae pneumonia (SMPP) in children using clinical data. Key predictors include dyspnea, CRP, and T lymphocytes, enabling early risk assessment.

Area of Science:

  • Pediatric Infectious Diseases
  • Machine Learning in Medicine
  • Computational Biology

Background:

  • Mycoplasma pneumoniae pneumonia (MPP) can lead to severe outcomes in children.
  • Early identification of severe cases is crucial for timely intervention.
  • Predictive models can aid in clinical decision-making.

Purpose of the Study:

  • To develop and validate an interpretable machine learning model for predicting severe Mycoplasma pneumoniae pneumonia (SMPP) in pediatric patients.
  • To identify key clinical and immunological predictors of SMPP.
  • To create a tool for early risk assessment.

Main Methods:

  • A Gradient Boosting Machine (GBM) model was developed using clinical and immunological data from 402 pediatric MPP patients.
  • Data was split into training (70%) and internal test (30%) sets.
  • 113 algorithms were evaluated, and SHapley Additive exPlanations (SHAP) were used for interpretability.

Main Results:

  • The GBM model achieved an AUC of 0.805 on the internal test set.
  • Sixteen core predictors were identified, including dyspnea, C-reactive protein (CRP), and T lymphocyte counts.
  • SHAP analysis highlighted dyspnea, elevated CRP, and decreased T lymphocytes as significant risk drivers.

Conclusions:

  • The developed GBM model demonstrates effectiveness in predicting SMPP risk.
  • The model's interpretability facilitates understanding of disease drivers.
  • Future integration into electronic medical record (EMR) systems is planned for real-time risk assessment.
Abstract

Keywords:
Mycoplasmachildmachine learningpneumoniaerisk assessment

Related Concept Videos

Atypical Pneumonia01:14

Atypical Pneumonia

Atypical pneumonia, often caused by Mycoplasma pneumoniae, is a form of pulmonary infection that differs from the classical presentation of bacterial pneumonia in both its cause and clinical symptoms. Mycoplasma pneumoniae is a pleomorphic bacterium notable for its lack of a rigid cell wall. This structural characteristic imparts resistance to beta-lactam antibiotics and significantly influences the bacterium’s behavior within the human host.Other pathogens responsible for the disease include...
Pneumonia III: Complications and Assessment01:30

Pneumonia III: Complications and Assessment

Pneumonia poses the potential for numerous complications that warrant consideration. These complications include the following:
Pneumonia I: Introduction01:29

Pneumonia I: Introduction

Pneumonia is an infection of the lower respiratory tract that leads to inflammation of the lung parenchyma, often resulting in the accumulation of inflammatory exudate in the alveoli and airways. Unlike the watery, low-protein fluid exudate in pulmonary edema, the exudate in this case is a thick fluid rich in immune cells, proteins, and debris produced during infection and inflammation.This impairs gas exchange and can lead to consolidation of lung tissue. The infection may be caused by a...