GA-EMDNet: Graph Attention-Guided Eye Movement Detection Network From EMG Sensor Signals

Abstract

Investigating and categorizing muscle-generated bioelectrical activity specifically electromyography (EMG) recordings associated with extraocular muscles (EOM) is fundamental for building advanced assistive systems. The dynamic and time-varying nature of these physiological waveforms demands analytical techniques capable of capturing temporal dependencies for accurate interpretation and classification. In this study, we propose a compact graph-based representation for EMG of EOM signal classification that transforms 1-D temporal waveforms into structured relational graphs. Specifically, each signal instance is segmented into temporal windows that are modeled as graph nodes, with edges encoding both local temporal continuity and feature-level similarity, enabling the learning of nonlocal dependencies. A graph attention network is employed to adaptively weight internode relationships and extract discriminative representations without reliance on frequency-domain decomposition or signal reconstruction, replacing complex domain-specific feature engineering with lightweight statistical descriptors computed directly on raw temporal windows. The proposed approach achieves a peak classification accuracy of 99.17% and mean classification accuracy of 98.33% (95% CI: [96.79%, 99.55%]) across a fivefold cross-validation with 80–20 split and 98.17% (95% CI: [96.99%, 99.35%]) across a tenfold cross-validation with 90–10 split, demonstrating competitive performance against strong classical baselines. The results highlight the importance of appropriate graph construction in leveraging attention-based graph models for time-dependent biomedical signal analysis.