Respiratory failure & ventilation
Deep learning classification of capnography waveforms: secondary analysis of the PRODIGY study
Authors' conclusion
The neural network could reduce false capnography alarms.
Study context
- Population
- Random sample of participants from 9 North American sites of the PRODIGY study10,391 waveform segments
- Participants
- 400 participants
- Primary endpoint
- Accuracy of deep learning classification of capnography waveform segments as 'breath' or 'no breath'
Methodological detailsdiagnostic accuracy study, post hoc, multicentre, internal-external validation
- Diagnostic accuracy study
- Post hoc
- Multicentre
- Internal-external validation
Primary result
As reported in the abstract
| Group | Accuracy | Precision | Recall |
|---|---|---|---|
| Neural network | 0.97 | 0.97 | 0.96 |
Neural network
Accuracy0.97
Precision0.97
Recall0.96
Further findings
Performance across hospitals (internal-external validation)
- Performance was consistent across hospitals
Original publication
Deep learning classification of capnography waveforms: secondary analysis of the PRODIGY study. J Clin Monit Comput. Published 13 May 2023. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →