Sustainability and predictive accuracy evaluation of gel and embroidered electrodes for ECG monitoring

Despite standards on electrocardiogram (ECG) monitoring in medical diagnostics, signal acquisition is prone to noisy artifacts and relies greatly on the quality of skin contact and signal transducing interference. Electrodes, serving as indispensable conduits in ECG signal acquisition, act as the crucial interface between the human body and recording instrumentation. The usage of traditional gel electrodes may provoke skin irritation, by contrast, the advent of embroidered electrodes, a contemporary innovation, holds the promise of more comfort monitoring. However, the sustainability quotient and predictive precision of these novel electrodes require more in-depth investigation. This paper endeavours a comprehensive evaluation of both gel and embroidered electrodes concerning ECG signal acquisition. Simultaneously, the study aims to construct a polynomial regression model leveraging advanced machine learning (ML) tools. The inferred model predicts embroidered signals based on experimental data obtained from gel electrodes. The methodology encompasses systematic data collection, preprocessing, insightful analysis, and the application of data-driven techniques. The findings of this study highlight the viability of embroidered electrodes as a compelling alternative, transcending traditional paradigms. Employing ML tools, the developed model achieves predictive accuracy, reflected in robust R2 values extending up to 94.9%.