Valvular & structural heart disease
AI-Enhanced Prediction of Aortic Stenosis Progression: Insights From the PROGRESSA Study
Authors' conclusion
This study demonstrates how an artificial intelligence-guided approach in clinical routine could help enhance risk stratification of AS.
Study context
- Population
- Patients with mild-to-moderate AS at baseline in the PROGRESSA study
- Participants
- 303 patients
- Comparison
- Machine/deep learning models vs standard clinical logistic regression model
- Outcome
- Prediction of AS progression (AUC)
- Follow-up
- Next 2 years and next 5 years
Methodological detailsmachine and deep learning, longitudinal annual follow-up
- Machine and deep learning
- Longitudinal annual follow-up
Performance
As reported in the abstract
Prediction of AS progression (average AUC across patient visits)
| Test or model | AUC at 2 years | AUC at 5 years |
|---|---|---|
| LightGBM | 0.85 | 0.83 |
LightGBM
AUC at 2 years0.85
AUC at 5 years0.83
Original publication
AI-Enhanced Prediction of Aortic Stenosis Progression: Insights From the PROGRESSA Study. JACC Adv. Published 11 September 2024. doi:10.
Summarised from the published abstract; every number is checked against it