EUVIMEDΑνακαλύψτε το EUVIMED
Open access

Cross-Silo Federated Learning for Predicting Successful Mechanical Ventilation Weaning Across 5 Intensive Care Unit Databases: Retrospective Database Analysis

Cross-Silo Federated Learning zur Vorhersage eines erfolgreichen Weanings von der mechanischen Beatmung über 5 Intensivstationsdatenbanken: retrospektive Datenbankanalyse

Seyedmostafa Sheikhalishahi iD, Mathias Kaspar iD, Johanna Schwinn iD, Matthaeus Morhart iD, Philipp Simon iD, Ludwig Christian Hinske iD

JMIR Medical Informatics · 2026-08-14 · 14 · e79713

DOI: 10.2196/79713 ↗

Auf einen Blick

In 5 ICU-Datenbanken wurde geprüft, wie gut federiertes Lernen ein erfolgreiches Weaning von der mechanischen Beatmung vorhersagt. FL erreichte AUROC 0,74 und lag unter dem zentralen Lernen.

Fragestellung

Wie gut sagt federiertes Lernen ein erfolgreiches Weaning von der mechanischen Beatmung über 5 ICU-Datenbanken voraus, im Vergleich zu lokalem und zentralem Lernen?

Methode

Retrospektive Analyse von 5 ICU-Datenbanken; Vergleich von FL, LL und CL mit XGBoost; Bewertung mit AUROC, AUPRC, Precision, Recall und F1-Score.

Ergebnisse

24.521 Patient:innen. CL: AUROC 0,81, AUPRC 0,57, F1 0,54. FL: AUROC 0,74, AUPRC 0,56, F1 0,52. LL: AUROC 0,68-0,84, AUPRC 0,51-0,68, F1 0,49-0,67.

Einschränkungen

Privatsphäre wurde nicht formal gemessen; die Ergebnisse sind daher ein Leistungsvergleich und keine Privacy-Performance-Abwägung.

Schlussfolgerung der Autor:innen

Die Leistungen unterschieden sich zwischen den Lernansätzen. LL war innerhalb der jeweiligen Institution am besten, CL auf gepoolten Testdaten, FL zeigte geringere, aber akzeptable Leistung ohne direkten Austausch von Patient:innendaten.

Kernaussagen

  • FL: AUROC 0,74; AUPRC 0,56; F1 0,52
  • CL auf gepoolten Testdaten am besten: AUROC 0,81
  • LL variierte je nach Datenbank: AUROC 0,68-0,84

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

At a glance

Journal article in JMIR Medical Informatics, published online on 2026-08-14. Written by Seyedmostafa Sheikhalishahi and 5 co-authors. Institution: Digital Medicine, University Hospital Augsburg, Augsburg, Germany. Major MEDLINE topics: Federated Learning, Ventilator Weaning, Intensive Care Units, Respiration, Artificial. Cited 1 times so far (Crossref); 38 references. The full text is freely available in PubMed Central.

Auf einen BlickΠερίληψηSubjectsΒιβλιογραφικά στοιχείαCiteRelated studies

Abstract (original language)

Background

The prediction of weaning from mechanical ventilation (MV) can support clinical decision-making and help reduce the risk of weaning failure in intensive care units (ICUs). Cross-silo federated learning (FL) offers a promising approach to developing robust predictive models across multiple institutions without requiring the sharing of patient-level data.

Objective

This study aimed to evaluate the feasibility and efficacy of FL for predicting successful weaning from MV across 5 diverse ICU databases and to compare its performance with local learning (LL) and centralized learning (CL) approaches that differ in their data-sharing requirements.

Methods

We conducted a retrospective analysis using 5 ICU databases, namely the eICU Collaborative Research Database (eICU-CRD), Medical Information Mart for Intensive Care IV (MIMIC-IV), Universitätsklinikum Augsburg (UKA), High-Resolution ICU Dataset (HiRID), and Amsterdam University Medical Centers (AUMC), transforming clinical variables into the Observational Medical Outcomes Partnership (OMOP) Common Data Model. We defined successful weaning as a sustained reduction in positive end-expiratory pressure. We compared 3 learning approaches, FL, LL, and CL, using extreme gradient boosting (XGBoost). Performance was evaluated using the area under the receiver operating characteristic curve (AUROC), area under the precision-recall curve (AUPRC), precision, recall, and F1-score. All data use complied with local ethical regulations and institutional review board approvals, and all databases contained deidentified patient data accessed under institutional data use agreements and governed by applicable privacy regulations.

Results

A total of 24,521 patients were included across 5 databases. The CL model achieved an AUROC of 0.81, AUPRC of 0.57, and F1-score of 0.54 on pooled test data. The FL model achieved a macroaveraged AUROC of 0.74, AUPRC of 0.56, and F1-score of 0.52. LL model performance varied across databases (AUROC=0.68-0.84, AUPRC=0.51-0.68, F1-score=0.49-0.67), reflecting differences in data distribution and class balance.

Conclusions

Our findings highlight performance differences between learning approaches for MV weaning prediction. LL models achieved the highest performance within their respective institutions (AUROC=0.68-0.84), CL achieved the highest performance on pooled test data (AUROC=0.81), and FL showed lower but reasonable performance while avoiding direct sharing of patient-level data (AUROC=0.74). The choice between approaches depends on institutional data-sharing constraints, local dataset characteristics, and acceptable performance thresholds. As privacy was not formally measured, these findings should be read as a performance comparison rather than a privacy-performance trade-off.

Subjects

MeSH terms (MEDLINE)

Federated LearningVentilator WeaningIntensive Care UnitsRespiration, ArtificialRetrospective StudiesHumansDatabases, FactualROC Curve

Author keywords

AUMCHiRIDMIMIC-IVUKAcross-silo federated learningeICU-CRDintensive care unitmachine learningmechanical ventilationweaning

Authors and institutions

  • Digital Medicine, University Hospital Augsburg, Augsburg, Germany.
  • Anesthesiology and Operative Intensive Care, Faculty of Medicine, University of Augsburg, Augsburg, Germany.

ORCID profiles: Seyedmostafa Sheikhalishahi, Mathias Kaspar, Johanna Schwinn, Matthaeus Morhart, Philipp Simon, Ludwig Christian Hinske

Cite

APA

Sheikhalishahi, S., Kaspar, M., Schwinn, J., Morhart, M., Simon, P., & Hinske, L. C. (2026). Cross-Silo Federated Learning for Predicting Successful Mechanical Ventilation Weaning Across 5 Intensive Care Unit Databases: Retrospective Database Analysis. JMIR Medical Informatics, 14, e79713. https://doi.org/10.2196/79713

Vancouver

Sheikhalishahi S, Kaspar M, Schwinn J, Morhart M, Simon P, Hinske LC. Cross-Silo Federated Learning for Predicting Successful Mechanical Ventilation Weaning Across 5 Intensive Care Unit Databases: Retrospective Database Analysis. JMIR Medical Informatics. 2026;14:e79713. doi:10.2196/79713

BibTeX
@article{Sheikhalishahi2026,
  author = {Sheikhalishahi, Seyedmostafa and Kaspar, Mathias and Schwinn, Johanna and Morhart, Matthaeus and Simon, Philipp and Hinske, Ludwig Christian},
  title = {Cross-Silo Federated Learning for Predicting Successful Mechanical Ventilation Weaning Across 5 Intensive Care Unit Databases: Retrospective Database Analysis},
  journal = {JMIR Medical Informatics},
  year = {2026},
  volume = {14},
  pages = {e79713},
  doi = {10.2196/79713}
}
Start for free →