Arrhythmias & electrophysiology
Improved diagnostic performance of insertable cardiac monitors by an artificial intelligence-based algorithm
Authors' conclusion
The new AI-based ILR-ECG-A lowered the rate of FP ICM diagnoses significantly while retaining a > 98% sensitivity.
Study context
- Population
- Recipients of insertable cardiac monitors (ICM) followed by the Implicity monitoring platform
- Participants
- 546 recipients · 879 episodes
- Comparison
- Artificial intelligence (AI)-based ILR-ECG Analyzer (ILR-ECG-A) reclassification vs ICM diagnoses, against adjudication committee
- Outcome
- Diagnostic specificity and sensitivity of ILR-ECG-A vs adjudication committee (AC) adjudication
Methodological detailsOne episode per ICM abnormal diagnosis per patient, to avoid clusterization
- One episode per ICM abnormal diagnosis per patient, to avoid clusterization
Performance
As reported in the abstract
Diagnostic performance of ILR-ECG-A vs AC adjudication
| Test or model | Episodes classified abnormal | Adjudicated artefacts | Adjudicated false positive (FP) | Adjudicated abnormal |
|---|---|---|---|---|
| ICM | 879 | 80 (9.1%) | 283 (32.2%) | 516 (58.7%) |
ICM
Episodes classified abnormal879
Adjudicated artefacts80 (9.1%)
Adjudicated false positive (FP)283 (32.2%)
Adjudicated abnormal516 (58.7%)
Parts of this Brief are withheld: the automatic checks could not confirm them against the abstract.
Original publication
Improved diagnostic performance of insertable cardiac monitors by an artificial intelligence-based algorithm. Europace. Published 3 January 2024. doi:10.
Summarised from the published abstract; every number is checked against it