Assessment of Muscle Fatigue Using Phase Entropy of sEMG Signals During Dynamic Contractions of Biceps Brachii

Complexity measures of the information contained within Surface Electromyography (sEMG) signals can aid in monitoring progression of muscle fatigue during the performance of everyday activities. In this work, the complexity of the sEMG signal dynamics during repetitive task has been assessed using phase entropy. sEMG signals have been recorded from the belly of the biceps brachii muscle of the participants while performing the dynamic biceps curl exercise until fatigue induced failure. From the pre-processed signals, the non-fatigue and fatigue segments are extracted. The phase entropy measures are computed for varied choice of number of angular sections (
$K$
). The results demonstrate existence of considerable rate of variability in the sEMG signals in both non-fatigue and fatigue states. The phase entropy is observed to decrease in the fatigued condition of the muscle irrespective of the value of
$\boldsymbol{K}$
. The optimal choice of
$K$
is found to be 4 and 7 for which a mean decrease of ~4% in the phase entropy values is noted. From the results, it can be inferred that phase entropy can be utilized to assess the impact of muscle fatigue on the complexity of sEMG signals. Thus, this feature can be useful for efficient control and design of rehabilitative techniques.