Arrhythmias & electrophysiology
Deep Learning Can Unmask Conduction Tissue Disease From an Ambulatory ECG
Authors' conclusion
A deep learning-enabled ambulatory ECG is capable of unmasking underlying conduction tissue system disease. This tool may help identify patients with significant intermittent bradyarrhythmia, potentially improving timely diagnosis and management.
Study context
- Population
- Unselected patients with 14-day single-lead ambulatory ECG recordingsunselected 14-day ambulatory ECG recordings (mean age, 60.5±17.8 years, 60% female) · training 189,414 · tuning 45,982 · internal validation 43,390 · external validation 42,173
- Participants
- 320,959 patients
- Index test or model
- Deep learning model on the last 24 hours of each recording, free of bradyarrhythmias, against bradyarrhythmias from the previous 13 days
- Target condition
- Prior asystole from sinus arrest or complete heart block (daytime sinus pause of ≥3 s, anytime sinus pause of ≥6 s, complete heart block, or a composite of these bradyarrhythmias)
Methodological details24-hour, single-lead ECG, unselected 14-day single-lead ambulatory ECG recordings
- 24-hour, single-lead ECG
- Unselected 14-day single-lead ambulatory ECG recordings
Performance
As reported in the abstract
- AUROC, prior daytime sinus pause ≥3 s
- 0.89
- AUROC, anytime sinus pause ≥6 s
- 0.87
- AUROC, complete heart block
- 0.93
- AUROC, composite end point
- 0.89
- Negative predictive values
- between 97.9 and 99.9%
Further findings
Bradyarrhythmias within the following 13 days (composite end point), using the first 24 hours of ECG data
- AUC, composite end point
- 0.88
Original publication
Deep Learning Can Unmask Conduction Tissue Disease From an Ambulatory ECG. Circ Arrhythm Electrophysiol. Published 26 August 2025. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →