Hierarchical fusion of electrocardiogram and phonocardiogram data improves heart failure detection.

Ma, Fei; Zhang, Haobo; Fang, Siyi; Wang, Qimei; Lin, Fan; Chao, Lianying; Wang, Zhiwei; Li, Qiang et al. · Patterns (N Y) · 2026

Where this comes from

Abstract

Heart failure with reduced ejection fraction (HFrEF) is a clinical syndrome with high morbidity and mortality. Developing cost-effective and easily deployable screening methods is crucial for improving early diagnosis and management. We propose LGC<sup>2</sup>-Net, a hierarchical fusion network that leverages multi-channel electrocardiograms (ECGs) and phonocardiograms (PCGs) for HFrEF detection. LGC<sup>2</sup>-Net simultaneously exploits complementarity across modalities and channels with its channel-specific and channel-shared branches. Each channel-specific branch employs a local-to-global hierarchical attention mechanism to capture both local and global semantic information within each ECG-PCG pair. The channel-shared branch further aligns and aggregates features from all channels, enabling effective modeling of inter-channel correlations. We established a new multi-channel ECG-PCG dataset with 2,480 synchronized recordings collected from 620 subjects using a digital stethoscope. Experiments demonstrate that LGC<sup>2</sup>-Net surpasses existing methods by 7.42% in average accuracy, highlighting its potential as an accurate, non-invasive, and scalable tool for HFrEF screening.