EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Machine learning models for predicting plastic bronchitis in hospitalized children with Mycoplasma pneumoniae pneumonia undergoing clinically indicated bronchoscopy: a single-center retrospective study

Machine-Learning-Modelle zur Vorhersage von Plastic Bronchitis bei hospitalisierten Kindern mit Mycoplasma-pneumoniae-Pneumonie unter klinisch indizierter Bronchoskopie: eine monozentrische retrospektive Studie

Shuai Yu, Yifei Li, Qing Liu, Zhengfeng Xue, Yu Gao, Yuzhi Yang, Yuanxia Li

Frontiers in Pediatrics · 30.10.2026 · 14

DOI: 10.3389/fped.2026.1983032 ↗

Auf einen Blick

Bei 702 hospitalisierten Kindern mit Mycoplasma-pneumoniae-Pneumonie zeigten mehrere Machine-Learning-Modelle eine gute Vorhersage für Plastic Bronchitis. RF und CatBoost hatten die besten bzw. komplementären Leistungen.

Fragestellung

Welche Machine-Learning-Modelle sagen das Risiko für Plastic Bronchitis bei Kindern mit Mycoplasma-pneumoniae-Pneumonie voraus?

Methode

Retrospektive monozentrische Studie mit 702 Kindern; 7:3-Aufteilung in Trainings- und Testset; 8 Machine-Learning-Modelle; 5-fold stratified cross-validation, SHAP, temporale Validierung und Bootstrap-Korrektur.

Ergebnisse

9 Prädiktoren wurden identifiziert. Im Testset lag die ROC-AUC bei 0,854 bis 0,881. RF: Accuracy 0,853, Precision 0,804, Recall 0,721, F1 0,759, ROC-AUC 0,881. CatBoost: Accuracy 0,853, Precision 0,793, Recall 0,736, F1 0,762, ROC-AUC 0,871.

Einschränkungen

Monozentrische retrospektive Studie; laut Autor:innen ist eine multizentrische externe Validierung erforderlich.

Schlussfolgerung der Autor:innen

Modelle auf Basis klinischer, laborchemischer und bildgebender Merkmale zeigten Potenzial zur Risikoklassifikation. RF und CatBoost hatten komplementäre Stärken; SHAP ermöglichte die individuelle Interpretation.

Kernaussagen

  • 702 Kinder mit MPP, davon 225 mit PB
  • Testset-ROC-AUC: 0,854 bis 0,881
  • RF und CatBoost wurden als Kandidatenmodelle behalten

Deutsche Zusammenfassung: KI-gestützt aus dem Original-Abstract erstellt; alle Zahlen wurden automatisch mit dem Original abgeglichen. Maßgeblich ist der Original-Abstract.

Kurz gefasst

Fachartikel in der Zeitschrift Frontiers in Pediatrics, online erschienen am 30.10.2026. Verfasst von Shuai Yu und 6 weiteren Autor:innen. Beteiligte Einrichtung: Department of Pediatric Respiratory Medicine, Affiliated Hospital of Yan'an University, Yan'an, Shaanxi, China. Die Arbeit nennt 37 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

Auf einen BlickAbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Objective

To develop and compare multiple machine learning models for predicting the risk of plastic bronchitis (PB) in children with Mycoplasma pneumoniae pneumonia (MPP) and to establish an interpretable early risk identification tool.

Methods

A retrospective study was conducted involving 702 hospitalized children with MPP (477 non-PB and 225 PB) admitted to the Affiliated Hospital of Yan'an University from January 2023 to January 2026. Clinical characteristics, laboratory findings, and imaging data were collected. Key predictors were identified using a two-stage Lasso–Boruta approach. The dataset was randomly divided into training and test sets at a ratio of 7:3. Eight machine learning models, including CatBoost, XGBoost, LightGBM, Random Forest, Logistic Regression, Support Vector Machine, Gaussian Naive Bayes, and K-Nearest Neighbors, were developed. Hyperparameters were optimized using five-fold stratified cross-validation with grid search or random search based on the F 1 score. Model performance was evaluated using accuracy, precision, recall, F 1 score, ROC-AUC, PR-AUC, calibration, and decision curve analysis. SHAP analysis was applied for model interpretation. Temporal validation and bootstrap optimism correction were further performed to assess temporal generalizability and optimism in model performance.

Results

The Lasso–Boruta approach identified nine PB-related predictors, including fever duration, peak temperature, WBC, CRP, ALB, LDH, FDP, D-dimer, and pulmonary consolidation. All eight models demonstrated predictive capability in the test set, with ROC-AUC values ranging from 0.854 to 0.881. RF achieved an accuracy of 0.853, precision of 0.804, recall of 0.721, F 1 score of 0.759, and ROC-AUC of 0.881, whereas CatBoost achieved an accuracy of 0.853, precision of 0.793, recall of 0.736, F 1 score of 0.762, and ROC-AUC of 0.871. In temporal validation, RF achieved the highest accuracy (0.925) and F 1 score (0.900), whereas CatBoost achieved the highest ROC-AUC (0.970) and PR-AUC (0.956). Considering the complementary performance profiles across different evaluation dimensions and the results of bootstrap optimism correction, RF and CatBoost were retained as candidate models for further interpretation and risk prediction tool development. SHAP analysis provided individual-level explanations of the contributions of the selected predictors to PB risk.

Conclusions

Machine learning models based on clinical manifestations, laboratory indicators, and imaging features showed potential for PB risk identification in children with MPP. RF and CatBoost exhibited complementary predictive strengths across different evaluation dimensions, while SHAP analysis facilitated individual-level interpretation. Further multicenter external validation is required.

Schlagwörter

Stichwörter der Autor:innen

Mycoplasma pneumoniae pneumoniachildrenmachine learningplastic bronchitisrisk prediction model

Autor:innen und Institutionen

  • Department of Pediatric Respiratory Medicine, Affiliated Hospital of Yan'an University, Yan'an, Shaanxi, China.
  • Department of Pediatric Cardiology, West China Second University Hospital, Sichuan University, Chengdu, China.

Zitieren

APA

Yu, S., Li, Y., Liu, Q., Xue, Z., Gao, Y., Yang, Y., & Li, Y. (2026). Machine learning models for predicting plastic bronchitis in hospitalized children with Mycoplasma pneumoniae pneumonia undergoing clinically indicated bronchoscopy: a single-center retrospective study. Frontiers in Pediatrics, 14. https://doi.org/10.3389/fped.2026.1983032

Vancouver

Yu S, Li Y, Liu Q, Xue Z, Gao Y, Yang Y, et al. Machine learning models for predicting plastic bronchitis in hospitalized children with Mycoplasma pneumoniae pneumonia undergoing clinically indicated bronchoscopy: a single-center retrospective study. Frontiers in Pediatrics. 2026;14. doi:10.3389/fped.2026.1983032

BibTeX
@article{Yu2026,
  author = {Yu, Shuai and Li, Yifei and Liu, Qing and Xue, Zhengfeng and Gao, Yu and Yang, Yuzhi and Li, Yuanxia},
  title = {Machine learning models for predicting plastic bronchitis in hospitalized children with Mycoplasma pneumoniae pneumonia undergoing clinically indicated bronchoscopy: a single-center retrospective study},
  journal = {Frontiers in Pediatrics},
  year = {2026},
  volume = {14},
  doi = {10.3389/fped.2026.1983032}
}
Kostenlos starten →