An innovative adaptive filtering technique for electromyography (EMG) signals influenced by spinal cord transcutaneous stimulation

The study examines the interferences arising from scTS in electromyographic signals and investigates the effectiveness of the AA-IF technique in filtering out the articfacts. Surface electromyographic signals were obtained from Biceps Brachii (BB) and Triceps Brachii (TB) muscles in five spinal cord injury (SCI) patient groups who were exposed to scTS at the current amplitude ranging from 20 to 55 mA and the frequency ranging from 30 to 60 Hz. Analysing the results of Fast Fourier Transform (FFT) we determine peak amplitudes and frequency bands containing noise from scTS artifacts. We also used the AA-IF and the EMD-BF techniques to remove such artifacts from the resultant data points. To compare the efficiency of the described techniques, the authors used the preserved FFT content and the RMS of the EMG signal after enabling the given filter. The scTS artifacts influenced nearby frequency bins with about 2Hz width from the stimulator’s most significant frequency and its harmonics. The width of contaminated bands increased by the enhancing scTS current intensity (p < 0. 001), has been lesser during voluntary contraction in compared with rest condition (p < 0. 05) and the difference is significant in BB muscle in compared with TB muscle (p < 0. 001). By comparing these two merging strategies, it can be concluded that the AA-IF technique allowed to retain a greater part of the FFT compared to EMD-BF.