Motion Artifact Reduction Algorithm for Wearable Electrocardiogram Monitoring Systems

An electrocardiogram (ECG) test is usually performed through measurement with Ag/AgCl electrodes. However, this method has disadvantages, such as skin irritation, short shelf life, and conductance variation in gels over time, which limit the usability for long-term monitoring. One of the alternatives is adopting conductive fabric-based electrodes, but this method involves a problem in which a motion artifact (MA) becomes more severe because no gel and adhesive are available to provide a stable interface between the electrodes and the skin. To address this problem, this study presents a conductive fabric-based ECG monitoring system and an MA reduction algorithm. The system can simultaneously measure the ECG and the electrode–tissue impedance (ETI) with the proposed shirt. The MA reduction algorithm aims to reduce MA with ETI information. The MAs in Lead I, II, and III ECG measurement are generated by lifting arms, walking, and jogging. Results of the MA reduction are quantified by the correlation coefficient (CORR) and mean squared error (MSE). The quantitative analysis shows that the MA can be suppressed by the proposed algorithm. Moreover, compared with a reproduced existing approach, the proposed algorithm performs better in most cases.