TexSense: Comparative ECG Analysis of E-Textile and Sticky Electrodes During Motion Artifacts

The rapid advancement of wearable technology is transforming health monitoring and driving demand for more comfortable, long-term ECG solutions. Traditional sticky or gel electrodes (SE), while highly accurate in stationary medical environments, often cause discomfort and skin irritation with prolonged adhesive use. Textile-based dry electrodes (TE) eliminate the need for conductive gel, offering a promising alternative for continuous ECG monitoring in dynamic conditions. However, the performance of textile electrodes in real-world settings remains largely unexplored, highlighting a critical research gap in understanding how dry electrodes compare to sticky electrodes when motion artifacts are present. This study investigated a comparative evaluation of gel-based sticky and textile dry electrodes during exercise-induced motion artifact tasks. We performed the study with 10 adult healthy participants and designed 16 distinctive motion artifact tasks that included reclining, sitting, walking, and running. The qualitative, and quantitative analysis illustrated the comparison of ECG signal, Heart Rate (HR) detection accuracy and error%, signal quality, SNR, Short-Time Fourier Transform (STFT), Kurtosis, Entropy, Continuous Wavelet Transform (CWT), and Bland-Altman analysis. The analysis showed textile electrodes achieved an HR detection accuracy of approximately 99% and an average Signal-to-Noise Ratio (SNR) of 15.8 dB. This performance approaches the 17.3 dB average of sticky electrodes, demonstrating strong resilience to noise in low-to-moderate motion tasks. The STFT analysis indicated reliable capture of essential QRS frequency components (5–15 Hz) by textile electrodes, though high-motion tasks produced slight deviations that may limit detailed QRS morphology analysis. The CWT analysis revealed the TE showed an average magnitude difference of 0.15 in low-motion tasks to 0.5 in high-motion tasks compared to SE, indicating increased noise sensitivity above 30 Hz. The Bland-Altman analysis showed the mean differences between sticky and textile electrode readings below ±0.5 in low-motion and ±0.7 in high-motion tasks. The study highlights the potential prospects of textile-based dry electrodes for wearable electro-physiological monitoring applications in motion-induced environments.