Robust and Explainable Bicuspid Aortic Valve Diagnosis Using Stacked Ensembles on Echocardiography

TL;DR

This study presents an explainable AI ensemble model that accurately classifies bicuspid versus tricuspid aortic valves from routine echocardiography videos, supporting early detection in diverse clinical settings.

Contribution

The paper introduces a novel stacked ensemble approach with explainability features for BAV diagnosis using standard echocardiographic videos, achieving high accuracy and transparency.

Findings

Achieved an F1-score of 0.907 and recall of 0.877 in classification.

Utilized Grad-CAM and SHAP for model explainability and case-level auditability.

Demonstrated potential for reliable BAV detection in resource-limited settings.

Abstract

Transthoracic echocardiography (TTE) is the first-line imaging modality for diagnosing bicuspid aortic valve (BAV), yet diagnostic performance varies with operator expertise and image quality. We developed an explainable AI model that distinguishes BAV from tricuspid aortic valves (TAV) using routinely acquired parasternal long-axis (PLAX) cine loops. A multi-backbone video ensemble was trained and evaluated using a leakage-aware, stratified outer cross-validation protocol on $N{=}90$ patient studies (48 BAV, 42 TAV). Across fixed outer splits and 10 random seeds, the calibrated stacked ensemble achieved an outer-CV F1-score of $0.907$ and recall of $0.877$. Frame-level Grad-CAM localized salient evidence to the aortic root and leaflet plane, while globally aggregated SHAP values quantified each video backbone's contribution to the stacked prediction, enabling transparent, case-level auditability. These findings indicate that PLAX-based video ensembles can support reliable BAV/TAV classification from routine echocardiographic cine loops and may facilitate earlier detection in non-specialist or resource-limited clinical settings.