Respiratory failure & ventilation
Application progress of machine learning in patient-ventilator asynchrony during mechanical ventilation: a systematic review
Authors' conclusion
ML models show significant potential for identifying and predicting PVA, as evidenced by their high reported performance metrics.
Study context
- Population
- Adults receiving mechanical ventilation; original studies applying machine learning (ML) to assess patient-ventilator asynchrony (PVA)studies included of 74 articles
- Participants
- 14 adults
- Outcome
- ML performance for PVA identification
Methodological detailsPROSPERO-registered protocol, refined quality assessment method, Searched PubMed, Embase, Web of Science and Cochrane from inception to November 2024
- PROSPERO-registered protocol
- Refined quality assessment method
- Searched PubMed, Embase, Web of Science and Cochrane from inception to November 2024
Pooled result
As reported in the abstract
- F1 score range
- 0.731 to 0.988
- Distinct ML methods
- 11
- Targeted PVA types
- 19
- Studies on PVA identification
- 13
- Studies predicting PVA occurrence
- 1
- Studies with external validation
- 2
- Studies addressing class imbalance
- 5
- Studies describing hyperparameter tuning
- 4
Original publication
Application progress of machine learning in patient-ventilator asynchrony during mechanical ventilation: a systematic review. Crit Care. Published 10 July 2025. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →