Search papers, labs, and topics across Lattice.
The paper introduces an interpretable framework for detecting structural heart disease (SHD) from ECGs by integrating clinically meaningful ECG foundation-model predictors within a generalized additive model (GAM). This approach addresses the interpretability limitations of existing black-box deep learning models while maintaining strong predictive performance. Experiments on the EchoNext benchmark demonstrate improvements in AUROC, AUPRC, and F1 score compared to state-of-the-art deep learning baselines, along with interpretable insights into the relationships between ECG diagnoses and SHD risk.
By combining foundation models with interpretable statistical methods, this ECG analysis achieves state-of-the-art SHD detection while revealing how specific ECG features contribute to risk.
Structural heart disease (SHD) is a prevalent condition with many undiagnosed cases, and early detection is often limited by the high cost and accessibility constraints of echocardiography (ECHO). Recent studies show that artificial intelligence (AI)-based analysis of electrocardiograms (ECGs) can detect SHD, offering a scalable alternative. However, existing methods are fully black-box models, limiting interpretability and clinical adoption. To address these challenges, we propose an interpretable and effective framework that integrates clinically meaningful ECG foundation-model predictors within a generalized additive model, enabling transparent risk attribution while maintaining strong predictive performance. Using the EchoNext benchmark of over 80,000 ECG-ECHO pairs, the method demonstrates relative improvements of +0.98% in AUROC, +1.01% in AUPRC, and +1.41% in F1 score over the latest state-of-the-art deep-learning baseline, while achieving slightly better performance even with only 30% of the training data. Subgroup analyses confirm robust performance across heterogeneous populations, and the estimated entry-wise functions provide interpretable insights into the relationships between risks of traditional ECG diagnoses and SHD. This work illustrates a complementary paradigm between classical statistical modeling and modern AI, offering a pathway to interpretable, high-performing, and clinically actionable ECG-based SHD screening.
Tsinghua AI