Respiratory failure & ventilation
Flow starvation during square-flow assisted ventilation detected by supervised deep learning techniques
Authors' conclusion
Recurrent neural network model appears excellent to identify airway pressure deformation due to flow starvation.
Study context
- Population
- Adult critically ill patients under mechanical ventilation > 24 h on square-flow assisted ventilation6,428 breaths analysed
- Participants
- 28 participants
- Comparison
- Recurrent neural network vs convolutional neural network for detecting airway pressure deformation
- Outcome
- Accuracy in identifying airway pressure deformation (flow starvation)
Methodological detailsmulticentre, observational
- Multicentre
- Observational
Performance
As reported in the abstract
Accuracy
| Test or model | Accuracy |
|---|---|
| Recurrent neural network | 87.9% [87.6-88.3] |
| Convolutional neural network | 86.8% [86.6-87.4] |
Accuracy
Recurrent neural network87.9% [87.6-88.3]
Convolutional neural network86.8% [86.6-87.4]
Further findings
Airway pressure deformation severity (expert classification)
- Normal-mild
- 42%
- Moderate
- 23%
- Severe
- 34%
Double triggering
- Breaths with double triggering, always with severe airway pressure deformation
- 8.8%
Inspiratory effort in severe deformation (ΔPes subgroup)
- Severe deformation breaths with ΔPes > 10 cmH2O
- 74.4%
- Severe deformation breaths with ΔPes > 15 cmH2O
- 37.2%
Parts of this Brief are withheld: the automatic checks could not confirm them against the abstract.
Original publication
Flow starvation during square-flow assisted ventilation detected by supervised deep learning techniques. Crit Care. Published 14 March 2024. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →