Detecting Structural Heart Disease from Electrocardiograms via a Generalized Additive Model of Interpretable Foundation-Model Predictors

2026-03-03

AI Summary

This study addresses the limitations in early screening for structural heart disease (SHD)—namely, the high cost and limited accessibility of echocardiography and the lack of interpretability in existing AI models—by proposing a novel framework that integrates an interpretable ECG foundation model with a generalized additive model (GAM). For the first time, a pretrained ECG foundation model is embedded within a GAM architecture, delivering both high predictive performance and clinically interpretable risk attributions. Evaluated on the large-scale EchoNext ECG–echo dataset (>80,000 samples), the model outperforms current state-of-the-art deep learning approaches by 0.98% in AUROC, 1.01% in AUPRC, and 1.41% in F1 score, while achieving comparable performance using only 30% of the training data. Robust results across subgroups underscore its potential to advance synergistic innovation between interpretable AI and traditional statistical modeling.

Technology Category

• Machine Learning: Other Foundations of Machine Learning
• Humans and AI: Other Foundations of Human Computation & AI
• Natural Language Processing: Interpretability, Analysis, and Evaluation of NLP Models

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics

Abstract

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.

Problem

• Structural Heart Disease
• Electrocardiogram
• Interpretability
• Black-box Model
• Clinical Adoption

Innovation

• Generalized Additive Model
• Interpretable AI
• ECG Foundation Model
• Structural Heart Disease
• Transparent Risk Attribution