Respiratory failure & ventilation
Reverse triggering neural network and rules-based automated detection in acute respiratory distress syndrome.
Authors' conclusion
Automated detection of RT demonstrated good performance, with the potential application of these programs for research and clinical care.
Study context
- Population
- Mechanically ventilated patients with ARDStraining 15, validation 13 patients
- Comparison
- Rules-based programs and neural network artificial intelligence (AI) program vs manual identification of reverse triggering
- Outcome
- Detection performance for RT: sensitivity, specificity, PPV and F1 score
Methodological detailsprospective observational, training and validation datasets
- Prospective observational
- Training and validation datasets
Performance
As reported in the abstract
Performance of automated RT detection
| Test or model | Specificity | F1 score |
|---|---|---|
| Rules-based | >95% | >75% |
| AI | >95% | >75% |
Specificity
Rules-based>95%
AI>95%
F1 score
Rules-based>75%
AI>75%
Further findings
- Reverse triggering (RT) with breath stacking
- 24.4%
- Mid-cycle RT
- 37.8%
- Most common phenotypes
Original publication
Reverse triggering neural network and rules-based automated detection in acute respiratory distress syndrome. J Crit Care. Published 24 January 2023. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →