Amplitude threshold criteria improve surface electrode specificity during walking and functional movements

Contamination of electromyographic (EMG) data due to crosstalk in recordings from surface electrodes can lead to misinterpretation of results. The purpose of this study was to determine if removing a portion of the EMG signal normalized to a maximum voluntary contraction (MVC) would improve the specificity of surface electrode recordings. We hypothesized that setting an amplitude threshold to define when a muscle was active would remove that part of the myoelectric signal most likely to include crosstalk, without affecting the intensity or the onset and cessation times. Surface and intramuscular electrodes recorded signals from the same muscles of adults performing cyclic ankle movements and walking at self-selected speeds. Signals identified as crosstalk were eliminated when 15% and 18% of the amplitude of the normalized signal was removed and muscle timing or intensity was not significantly changed in most cases.