Hear the Heartbeat in Phases: Physiologically Grounded Phase-Aware ECG Biometrics

TL;DR

This paper tackles ECG-based identity verification by addressing the phase heterogeneity within cardiac cycles. It introduces the Hierarchical Phase-Aware Fusion (HPAF) framework, featuring Cardiac Phase Segmentation, Intra-Phase Representation with dual-branch MVFEs, Phase-Grouped Hierarchical Fusion, Global Representation Fusion, and a contrastive loss, complemented by a Heartbeat-Aware Multi-prototype (HAM) enrollment strategy. By preserving phase-specific morphology and variation cues and structurally fusing information, HPAF achieves state-of-the-art results on three public datasets in both closed-set and open-set settings, demonstrating improved robustness to heartbeat variability and noise. The approach offers a practical advance for physiologically grounded, liveness-aware ECG biometrics, with potential for deployment in continuous authentication systems, while future work will address R-peak localization reliability and R-peak-free phase segmentation.

Abstract

Electrocardiography (ECG) is adopted for identity authentication in wearable devices due to its individual-specific characteristics and inherent liveness. However, existing methods often treat heartbeats as homogeneous signals, overlooking the phase-specific characteristics within the cardiac cycle. To address this, we propose a Hierarchical Phase-Aware Fusion~(HPAF) framework that explicitly avoids cross-feature entanglement through a three-stage design. In the first stage, Intra-Phase Representation (IPR) independently extracts representations for each cardiac phase, ensuring that phase-specific morphological and variation cues are preserved without interference from other phases. In the second stage, Phase-Grouped Hierarchical Fusion (PGHF) aggregates physiologically related phases in a structured manner, enabling reliable integration of complementary phase information. In the final stage, Global Representation Fusion (GRF) further combines the grouped representations and adaptively balances their contributions to produce a unified and discriminative identity representation. Moreover, considering ECG signals are continuously acquired, multiple heartbeats can be collected for each individual. We propose a Heartbeat-Aware Multi-prototype (HAM) enrollment strategy, which constructs a multi-prototype gallery template set to reduce the impact of heartbeat-specific noise and variability. Extensive experiments on three public datasets demonstrate that HPAF achieves state-of-the-art results in the comparison with other methods under both closed and open-set settings.

Figures (5)

• Figure 1: An illustration of the relationship between the different phases of the cardiac cycle and the corresponding ECG signal.
• Figure 2: We present the proposed HPAF framework for phase-aware ECG identification. CPS detects R-peaks and partitions each heartbeat into four phase-aligned segments, namely P, QRS, ST, and T/U. Each phase is encoded by MVFE with MFEB and VFEB, and fused by PR-GAT to form phase representations, which PGHF and GRF then aggregate into a single beat-level identity embedding. The encoder is trained with the contrastive loss to increase inter-subject separability.
• Figure 3: CMC comparisons of different methods across three datasets. Subplots (\ref{['fig:cmc_closed_ecgid']})--(\ref{['fig:cmc_closed_ptb']}) denote the performance of different methods on the MIT-BIH, ECGID, and PTB datasets under the closed-set setting, respectively, while (\ref{['fig:cmc_open_ecgid']})--(\ref{['fig:cmc_open_ptb']}) show the corresponding results under the open-set setting.
• Figure 4: ROC comparisons of different variants. Subplots (\ref{['fig:roc_closed_ecgid']})– (\ref{['fig:roc_closed_ptb']}) show the ROC curves on ECGID, MIT-BIH, and PTB under the closed-set setting, respectively, whereas (\ref{['fig:roc_open_ecgid']})– (\ref{['fig:roc_open_ptb']}) depict the corresponding results under the open-set setting (same dataset order).
• Figure 5: Ablation study about the number of enrollment prototypes in HAM.