Stimulus Artifact Denoising Technology in Surface Electromyographic Signals Under Functional Electrical Stimulation

This paper presents a novel design to remove stimulus artifacts from electromyographic signals under functional electrical stimulation (FES). This design combines Kalman filter (KF), Comb filter (COMB), and Generalized Extreme Studentized Deviate (GESD) algorithms, and its effectiveness is verified by filtering synthetic semi-simulated and real data. Experimental results show that in the signal processed by the KF-COMB-G algorithm, the stimulus artifacts caused by FES are well suppressed. The filtered signal has a higher correlation coefficient and signal-to-noise ratio, which highlights the good effect of this design in removing FES interference. In addition, the algorithm is simple to implement and has strong real-time performance, providing a new and effective method for artifact removal processing of electromyographic signals under FES conditions.