Valvular & structural heart disease
Predictors of Paravalvular Leakage After Transcatheter Aortic Valve Replacement in Patients With Bicuspid Aortic Valve (BAV): A Machine Learning Model
Authors' conclusion
The machine learning model with 7 predictors effectively identifies post-TAVR PVL risk in patients with BAV stenosis, aiding procedural planning and PVL prevention. (Study on Standard Evaluation System and Optimal Treatment Path of Senile Valvular Heart Disease; NCT05044338).
Study context
- Population
- Patients with BAV undergoing transcatheter aortic valve replacement (TAVR)patients undergoing TAVR analyzed, 109 cases further verified the model · type 0 373 · type 1 707 · validation 109
- Participants
- 1,080 patients
- Comparison
- Random forest (RF) model and logistic model for predicting post-TAVR PVL
Methodological detailsrandom forest (RF) model, logistic regression analysis
- Random forest (RF) model
- Logistic regression analysis
Further findings
| Outcome | RF model | Logistic model |
|---|---|---|
| Model performance | ||
| AUC | 0.982 | 0.978 |
| Accuracy | 0.994 | 0.951 |
| Sensitivity | 0.997 | 0.917 |
| Specificity | 0.993 | 0.957 |
| AUC in the validation data set | AUC 0.975 | AUC 0.988 |
Predictors selected by the RF model
- 7 predictors: Corrected calcification volume (Corr-CV) of leaflets, Corr-CV of annular surroundings, Corr-CV of the left ventricular outflow tract, perimeter of the left ventricular outflow tract, plane/perimeter of the supra-annular plane, horizontal aorta, annular ellipticity, and postdilation
- 7
- Derivation data set
Decision curve analysis and clinical impact curve analysis
- Verified the clinical applicability and net benefit
Original publication
Predictors of Paravalvular Leakage After Transcatheter Aortic Valve Replacement in Patients With BAV: A Machine Learning Model. JACC Adv. Published 20 January 2026. doi:10.
Summarised from the published abstract; every number is checked against it