Towards the Development of Knitted Surface EMG Sensors: Impact of Knit Structures and Configuration on Monitoring Dynamic Muscle Contractions

This study investigates how knit stitch configurations affect the performance of textile-based surface electromyography (sEMG) electrodes for monitoring Biceps brachii activity. Four sensor variants were developed using plain, tuck, and warp stitches, with and without melting-yarn reinforcement. The evaluation combined controlled muscle contraction recordings, microscopic characterization, finite element simulations, and advanced signal analysis, including time-domain, frequency-domain, Short-Time Fourier Transform (STFT), Wavelet Packet Decomposition (WPD), Hilbert–Huang Transform (HHT), and bispectral methods. Results show that stitch geometry and yarn reinforcement directly influence electrode performance. Plain stitches with melting yarn produced high signal amplitudes but were more susceptible to noise and motion artifacts. Non-fused plain stitches showed reduced stability. Tuck stitches demonstrated lower deformation and noise floors; in particular, fused tuck designs balanced sensitivity and stability, providing short latency and high spectral fidelity. Warp stitches exhibited minimal static deformation and preserved higher-frequency content, although twisting during load introduced variability. Finite element results supported these trends by linking loop deformation ranges to contact stability at the electrode–skin interface. These findings establish that stitch topology and reinforcement are key determinants of signal quality in textile electrodes. The results provide practical design guidelines for tailoring textile sEMG sensors to application-specific needs, including fitness monitoring, rehabilitation, and prosthetic control.