Chaotic analysis of electromyography signal at low back and lower limb muscles during forward bending posture.

The present study aims to perform further investigation on muscular activity during forward bending posture by applying a nonlinear dynamic (chaotic) analysis method. The objective is to determine the characteristics of the low back and lower limb muscle electromyography (EMG) signal under chaotic analysis while maintaining a certain posture. Twelve subjects were asked to maintain postures of six stages in bending angles from 0 to 180 degrees, and the EMG signals of erector spinae (ES) at L1 and L5 levels, hamstrings, and gastrocnemius were recorded. Two important concepts to characterize deterministic chaos, Correlation Dimension (D2) and Lyapunov Exponents (lambda, LE), were applied to observe the chaotic characteristic of the EMG signals, and the results were also compared to the FFT based total power value. The EMG signals in all observed muscles during bending posture showed results of positive LE and high D2 at 5.5 to 7.5, which led us to classify EMG as a high dimension chaotic signal. The result obtained showed that the correlation dimension could be used as a reliable method to compare the EMG signal in various postures (or muscle contraction conditions). However, Lyapunov exponents did not show a significant difference of comparison result thus leading to the conclusion that LE could not be a reliable measure for high dimension chaotic system, such as an EMG signal. It was also shown that in both light and deep bending, the EMG signal of the low back muscles was of the same complexity level due to the D2 result. It was evident that somehow the low back muscle remained loaded in all bending stages which was contrary to the hypothesis that the low back muscle was less active during the deep bending, as was the case in most of the previous studies. The reason of such phenomenon was elucidated with use of the theory of muscular functional differentiation, including corticalization and spinalization.