DeepArrhythmia: Segment-Contextualized ECG Arrhythmia Classification via Selective Evidence Acquisition

2026-05-14

AI Summary

This study addresses a critical limitation in existing electrocardiogram (ECG)-based arrhythmia detection methods, which typically treat individual heartbeats as isolated samples and neglect the essential role of rhythmic context in accurate classification. To overcome this, the authors propose DeepArrhythmia, a novel framework that integrates raw ECG signals with waveform images to precisely localize R-peaks across multi-beat segments and jointly leverages both rhythmic and morphological features for structured, heartbeat-level predictions. The approach innovatively decouples physiological measurement from evidence integration and introduces a confidence-based dynamic routing mechanism that adaptively switches between minimal and rich evidence states. This design effectively balances contextual awareness with physiological interpretability, significantly improving classification accuracy, computational efficiency, and decision transparency.

Abstract

Beat-level Electrocardiography (ECG) arrhythmia detection aims to assign an arrhythmia class to each beat in a recording, yet many existing systems treat beats as isolated local instances. This is limiting because beat labels often depend on multi-beat rhythm context, including timing, compensatory pauses, and beat-to-beat morphological consistency. We present DeepArrhythmia, a tool-grounded multimodal framework for segment-contextualized beat-level ECG arrhythmia classification. Given a multi-beat ECG segment, DeepArrhythmia combines the raw ECG signal and a rendered waveform image, localizes R peaks to identify beat instances, and produces structured beat-level predictions. The framework decouples physiological measurement from evidence integration using specialized tools for beat localization, numerical rhythm--morphology extraction, and morphology-focused textual analysis. DeepArrhythmia uses segment-level confidence to route between minimal and rich evidence states, since richer physiological evidence is not uniformly useful. This agentic design integrates rhythm context, explicit physiological grounding, and selective evidence acquisition for decision making.

Problem

• ECG arrhythmia classification
• beat-level detection
• rhythm context
• morphological consistency
• compensatory pauses

Innovation

• segment-contextualized
• selective evidence acquisition
• multimodal ECG analysis
• beat-level arrhythmia classification
• rhythm context modeling