Valvular & structural heart diseaseCohort study
Comprehensive Aortic Stenosis Characterization Using Multiview Deep Learning
- Cohort study
- Trained on 210,193 images from 16,076 studies from KPNC
Authors' conclusion
EchoNet-AS synthesizes information from both B-mode videos and Doppler images to accurately assess AS severity. Its strong performance generalizes robustly to external validation cohorts and shows potential as an automated clinical decision support tool.
Study context
- Population
- Echocardiographic studies to assess AS severity16,076 training studies
- Outcome
- Discrimination of severe AS (AUC)
Main finding
As reported in the abstract
Discrimination of severe AS (EchoNet-AS)
AUC, KPNC held-out cohort
0.964 (95% CI: 0.952-0.973)
AUC, KPNC temporally distinct cohort
0.986 (0.983-0.989)
AUC, SHC external cohort
0.985 (0.975-0.992)
AUC, CSMC external cohort
0.989 (0.986-0.992)
Observational data: associations, not effects.
Original publication
Comprehensive Aortic Stenosis Characterization Using Multiview Deep Learning. JACC Cardiovasc Imaging. doi:10.1016/j.jcmg.2026.07.013
Summarised from the published abstract; every number is checked against it