Enhancement of Stress Electrocardiogram (ECG) Performance with Machine Learning: A Single-Center Study
Authors' conclusion
This study presents a proof of concept demonstrating that an AI-based model for stress ECG interpretation is feasible and shows acceptable performance.
Study context
- Population
- Patients who underwent coronary angiography within 2 months of exercise stress ECG
- Participants
- 1,200 ECGs
- Comparison
- Artificial intelligence (AI) model for exercise stress ECG interpretation
- Outcome
- Performance in predicting severe CAD
Methodological details5-fold cross-validation, external validation cohort
- 5-fold cross-validation
- External validation cohort
Performance
As reported in the abstract
Severe CAD, initial test subset
| Test or model | Sensitivity | Specificity | Accuracy |
|---|---|---|---|
| AI model | 93.6% | 93.2% | 93.4% |
AI model
Sensitivity93.6%
Specificity93.2%
Accuracy93.4%
Further findings
Severe coronary artery disease (CAD) prediction, second validation cohort (91 patients)
- Accuracy of the model in the second validation cohort
- 78%
- Sensitivity of the model in the second validation cohort
- 64.6%
- Specificity of the model in the second validation cohort
- 93%
Sensitivity gain over ESE, by sex (subgroup)
- Absolute increase in sensitivity, women
- 40.9%
- Absolute increase in sensitivity, men
- 44.6%
Original publication
Enhancement of Stress ECG Performance with Machine Learning: A Single-Center Study. JACC Adv. Published 11 September 2025. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →