Abstract Summary
Surface electromyograms (EMG) are susceptible to motion artifacts, especially during dynamic non-isometric contractions. A novel method is presented for detecting motion artifact contamination in EMG recordings. Specifically, root mean square (RMS) amplitudes of EMG recordings with two high-pass corner frequencies are computed (RMS1Hz and RMS20Hz) and the Spearman rank correlation coefficient, ρ, is calculated between RMS20Hz amplitudes and RMS20Hz/RMS1Hz amplitude ratios. For EMG recordings that are free of motion artifacts, RMS20Hz and RMS20Hz/RMS1Hz are strongly correlated (ρ~1), but for EMG recordings contaminated by large recurring motion artifacts, RMS20Hz and RMS20Hz/RMS1Hz are weakly or negatively correlated (ρ
Sign in to ORCID