Chaotic Complexity Determination of Surface EMG Signals

Neuromuscular system is highly nonlinear. Investigation of surface electromyographic (sEMG) signals during varied myoneural states can reveal nonlinear dynamical behavior of the system. This work aims to detect the chaotic nature of sEMG signals using various methods. Myoelectric signals are acquired from biceps brachii muscle of 45 participants using a standard protocol. The 0–1 and modified 0–1 tests are applied directly on the recorded signals. Permutation and bubble entropies (peEn, bEn) are also calculated for the recorded signals. The mean value of growth rate is found to be 0.998 for 0–1 test and 0.65 for modified 0–1 test. Mean values of peEn, bEn are obtained as 0.91 and 0.48, respectively. Chaotic complexity is indicated by 0–1 test and the entropy features, whereas stochastic nature is observed from the k value of modified 0–1 test. The computational complexity is less for 0–1 test compared to other methods. Hence, this method may be used for the determination of chaotic behavior of sEMG signals under various clinical conditions.