Study And Application Of Surface Myoelectric Signal Based On Time-Frequency And Time-Scale Analysis

Electromyography (EMG) is a technique for determining integrated functions of the neuromuscular system. Surface myoelectric signal (SMES) are generated by the quasi-random activation of the individual muscle fibers that make up a skeletal muscle, and reflects the anatomical and physiological properties of the muscle. The dsetection and processing of SMES permits the study of physio-electrical properties in the skeletal muscles during contraction. The understanding of electrical properties of muscle is essential for proper interpretations of SMES in the clinical domain. Analysis technique of SMES signal is currently widely used in biomechanic, ergonomic and kinesiologic studies for its convenience and non-invasiveness. SMES can also be a very valuable analytical method of normal and abnormal neuromuscular functioning if applied, processed and interpreted properly.Surface myoelectric signal can be used to analyse muscle activity from a global muscle perspective. A classic application of this technique for global muscle assessment is the evaluation of muscle fatigue. Muscle fatigue is generally defined as a transient decrease in the capacity of a muscle to perform work, resulting from prior physical activity. As the individual muscle fibers fatigue, more and more of those signals vanish, resulting in the frequency spectrum of a SMES shift to lower frequencies. Such changes are referred to as myoelectric manifestations of localized muscle fatigue. The spectrum change is usually studied by means of spectral parameters, such as the mean and median frequency of the power spectrum of the signal. By keeping track of myoelectric signal spectrum changes, we may obtain useful information on the physiological activity of muscles and nerves.Surface myoelectric signal analysis has been proved effective for assessing the electrical manifestations of localized muscle fatigue. In the past, the analysis of muscle fatigue has been restricted to isometric, constant force contractions due to the limitation of signal processing technique. During sustained static contractions SMES may be assumed to be quasi-stationary, that is stationary during short time intervals (0.5-2 s) and when the signal can be considered as a realization of a wide-sense stationary stochastic processs with Gaussian distribution of amplitude and zero mean. . Under this assumption spectral analysis based on the Fourier Transform may be applied. However, in many cases the assumption does not hold and othermethods must be used. This is always the case either for burst activation in isometric conditions or for dynamic contractions. Dynamic muscle contraction yield non-stationary myoelectric signals. The analysis of dynamic contractions, during which the surface myoelectric signal is a nonstationary stochastic process, is still missing.Recently, the availability of spectral estimation techniques specifically designed for non-stationary signal analysis made it possible to extend the employment of muscle fatigue assessment to dynamic contractions. This paper describes the time-frequency method to analyze the nonstationary surface myoelectric signals during dynamic contractions is introduced. The time-frequency distribution of the surface myoelectric signal was computed by means of STF^ Wigner-Ville distribution and Choi-Williams distribution. They are used for the visual inspection of the evolution of the frequency content of the signal. Quadratic time-frequency representations do not require stationarity and are being investigated in this study. Most of them introduce unacceptable artifacts due to cross-terms. The Choi-Williams transform is particularly promising because of the small amplitude of the cross-terms. In general, experimental data obtained during dynamic contractions should be regarded with caution because of the limited information available about the effect of the relative muscle-electrode movement. Changes in muscle length, force, and electrode position contribute to the nonstationarity of the surface myoelectric signal. These factors, unrelated to localized...