Interpretable Multimodal Zero-Shot ECG Diagnosis via Structured Clinical Knowledge Alignment

TL;DR

ZETA is a novel zero-shot multimodal framework that enhances ECG diagnosis interpretability by aligning signals with structured clinical observations, improving transparency and generalization without disease-specific training.

Contribution

The paper introduces ZETA, a zero-shot ECG diagnosis model that uses structured clinical knowledge and multimodal embedding alignment for improved interpretability and generalization.

Findings

Competitive zero-shot classification performance

Enhanced interpretability with clinically relevant features

Grounded predictions in positive and negative diagnostic evidence

Abstract

Electrocardiogram (ECG) interpretation is essential for cardiovascular disease diagnosis, but current automated systems often struggle with transparency and generalization to unseen conditions. To address this, we introduce ZETA, a zero-shot multimodal framework designed for interpretable ECG diagnosis aligned with clinical workflows. ZETA uniquely compares ECG signals against structured positive and negative clinical observations, which are curated through an LLM-assisted, expert-validated process, thereby mimicking differential diagnosis. Our approach leverages a pre-trained multimodal model to align ECG and text embeddings without disease-specific fine-tuning. Empirical evaluations demonstrate ZETA's competitive zero-shot classification performance and, importantly, provide qualitative and quantitative evidence of enhanced interpretability, grounding predictions in specific, clinically relevant positive and negative diagnostic features. ZETA underscores the potential of aligning ECG analysis with structured clinical knowledge for building more transparent, generalizable, and trustworthy AI diagnostic systems. We will release the curated observation dataset and code to facilitate future research.