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

In three linesDeepArrhythmia is a multimodal framework for beat-level ECG arrhythmia classification. It combines raw signal and waveform image, localizes R peaks, and uses specialized tools for rhythm and morphology extraction. The system selectively routes between minimal and rich evidence states based on segment-level confidence.

Summary generated by Claude — human-verified