Valvular & structural heart disease
Machine Learning Prediction for Prognosis of Patients With Aortic Stenosis
Authors' conclusion
In patients with AS, a machine learning algorithm predicts outcomes with good accuracy, and prognostic characteristics were identified.
Study context
- Population
- Patients with aortic stenosis (AS) undergoing echocardiography
- Participants
- 10,407 / 11,738 / 954
- Comparison
- Primary vs External cohort 1 vs External cohort 2
- Primaryn = 10,407
- External cohort 1n = 11,738
- External cohort 2n = 954
- Outcome
- Composite of:
- Aortic valve replacement
- Mortality
- Follow-up
- 1 and 5 years
Methodological detailsmachine learning, random survival forest, 2 external validation cohorts
- Machine learning
- Random survival forest
- 2 external validation cohorts
Performance
As reported in the abstract
Composite of:
- AVR
- Mortality (RSF prediction)
| Test or model | AUC at 1 year (primary cohort) | AUC at 5 years (primary cohort) | Sensitivity (cut-off 50%) | Specificity (cut-off 50%) | AUC external cohort 1 | AUC external cohort 2 |
|---|---|---|---|---|---|---|
| Random survival forest (RSF) | 0.83 (95% CI: 0.80-0.86) | 0.83 (95% CI: 0.81-0.84) | 0.80 | 0.73 | 0.73 (95% CI: 0.72-0.74) | 0.74 (95% CI: 0.72-0.76) |
Random survival forest (RSF)
AUC at 1 year (primary cohort)0.83 (95% CI: 0.80-0.86)
AUC at 5 years (primary cohort)0.83 (95% CI: 0.81-0.84)
Sensitivity (cut-off 50%)0.80
Specificity (cut-off 50%)0.73
AUC external cohort 10.73 (95% CI: 0.72-0.74)
AUC external cohort 20.74 (95% CI: 0.72-0.76)
Further findings
Events in primary cohort · median follow-up 48 (21-87) months
- Aortic valve replacement
- 1,116
- Died
- 5,069
Parts of this Brief are withheld: the automatic checks could not confirm them against the abstract.
Original publication
Machine Learning Prediction for Prognosis of Patients With Aortic Stenosis. JACC Adv. Published 14 August 2024. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →