Linlin Zhao, Yongxiang Zou, Xiuxiu Shi, Long Cheng. LiteMamba-EMG: A Lightweight Mamba-based Framework for Robust sEMG-based Swallowing Activity Classification in Dysphagia PatientsJ. Machine Intelligence Research, 2026, 23(6): 1398-1412. DOI: 10.1007/s11633-026-1663-8
Citation: Linlin Zhao, Yongxiang Zou, Xiuxiu Shi, Long Cheng. LiteMamba-EMG: A Lightweight Mamba-based Framework for Robust sEMG-based Swallowing Activity Classification in Dysphagia PatientsJ. Machine Intelligence Research, 2026, 23(6): 1398-1412. DOI: 10.1007/s11633-026-1663-8

LiteMamba-EMG: A Lightweight Mamba-based Framework for Robust sEMG-based Swallowing Activity Classification in Dysphagia Patients

  • Accurate swallowing assessment via surface electromyography (sEMG) is hindered by limited channel interaction modeling, uniform temporal processing of event-driven signals, and scarce labeled clinical data. We propose SwallowMamba, a hybrid framework that integrates multi-scale convolutional feature extraction with a multi-channel cooperative scanning state space model (MCCS-SSM). MCCS-SSM assigns channel-specific state space parameters and employs a time-varying dynamic cooperation matrix via scaled dot-product attention to model inter-muscle coordination at each time step. An event-aware adaptive step-size (EAAS) mechanism modulates temporal resolution by integrating short-time energy, spectral flux, and instantaneous amplitude variation, allocating finer discretization to swallowing events and coarser steps to rest. To address labeled data scarcity, a hierarchical VICReg (H-VICReg) self-supervised pretraining strategy leverages the large-scale NinaPro hand sEMG dataset for cross-anatomical transfer with multi-layer regularization. Evaluated on a five-class swallowing sEMG dataset (11 subjects, including 3 laryngeal cancer patients) using leave-one-subject-out cross-validation, SwallowMamba achieves 93.5% ± 0.3% accuracy with 2.56 M parameters and 5.2 ms inference latency, outperforming Transformer, MoEMba, and EEGMamba baselines. Ablation experiments confirm the individual and synergistic contributions of each module.
  • loading

Catalog

    /

    DownLoad:  Full-Size Img  PowerPoint
    Return
    Return