Arrhythmias & electrophysiologyDiagnostic or prognostic study
Artificial intelligence for direct-to-physician reporting of ambulatory electrocardiography
- Diagnostic accuracy study
Authors' conclusion
We conclude that the DeepRhythmAI model has excellent negative predictive value for critical arrhythmias, substantially reducing false-negative findings, but at a modest cost of increased false-positive findings.
Study context
- Population
- Ambulatory electrocardiogram (ECG) recordings (mean duration 14 ± 10 days)
- Participants
- 14,606 ECG recordings
- Comparison
- Technicians vs DeepRhythmAI
- Techniciansn = 167reference
- Outcome
- Sensitivity for identification of critical arrhythmias
Performance
Critical arrhythmias
Relative risk
14.1
95% CI 10.4–19.0
By group
DeepRhythmAI
98.6% (95% CI 97.7-99.4)
Mean sensitivity · False-negative findings 3.2/1,000 patients
Technicians
80.3% (95% CI 77.3-83.3%)
Mean sensitivity · False-negative findings 44.3/1,000 patients
Technicians vs DeepRhythmAI
RR 14.1 (95% CI 10.4–19.0)
As reported in the abstract
Safety
False-positive event rate
Original publication
Artificial intelligence for direct-to-physician reporting of ambulatory electrocardiography. Nat Med. Published 10 February 2025. doi:10.1038/s41591-025-03516-x
Summarised from the published abstract; every number is checked against it