Heart failure & cardiomyopathies
Artificial Intelligence for the Detection of Hypertrophic Cardiomyopathy From Standard Electrocardiogram
Authors' conclusion
The AI ECG algorithm demonstrated highly specific HCM detection confirmed by cMRI and may improve early identification of unrecognized HCM.
Study context
- Population
- Patients with cardiac magnetic resonance imaging (cMRI) for suspected hypertrophic cardiomyopathy (HCM), with confirmed HCM and analyzable electrocardiograms (ECGs), and control patients without cardiomyopathyall patients, of 314 patients with cMRI for suspected HCM, 150 had analyzable ECGs and confirmed HCM · HCM 150 · Control patients without cardiomyopathy 83
- Participants
- 233 patients
- Comparison
- Viz HCM AI ECG labels vs ground truth
- HCMn = 150
- Control patientsn = 83
Performance
As reported in the abstract
Diagnostic performance for HCM detection
Area under the curve
0.946
95% CI 0.916–0.970
Sensitivity
58%
95% CI 50.0%–65.6%
Specificity
100%
95% CI 95.6%–100%
Further findings
Predictor of correct detection: apical subtype
| Comparison | Adjusted OR (95% CI) | P value | |
|---|---|---|---|
| Apical subtype · multivariable logistic regression | 4.71 (1.71–15.48) | P = 0.005 | |
| 0.050.21520Adjusted odds ratio (log scale) |
Early detection before clinical diagnosis
- Patients with ECGs available at least 1 year prior to cMRI in whom HCM was detected
- 9 of 28
- Time before clinical diagnosis (median), among those 9
- 2.6 years
Baseline characteristics
- Mean age of all 233 patients
- 56 years
- Male
- 62%
Original publication
Artificial Intelligence for the Detection of Hypertrophic Cardiomyopathy From Standard Electrocardiogram. JACC Adv. Published 16 June 2026. doi:10.
Summarised from the published abstract; every number is checked against it