M.M. Mezhennaya, A.N. Osipov, I.A. Ilyasevich, N.S. Davydova, M.V. Davydov, V.A. Kulchitsky
Global electromyography is a recognized method studying of neuromuscular system, which is founded on the registration and the qualitative-quantitative analysis of global bioelectrical activity of motor unit population by means of epicutaneous electrodes. Electromyogram (EMG) parameters are used as an objective diagnostic indicator of muscle group functional condition.
Traditional methods of global electromyogram analysis are limited to the calculation statistical parameters (maxi-mum and average arithmetical amplitude, average arithmetical frequency potentials) and the visual estimation of general type EMG-signal in according to J.S.Jusevich's classification. Thus EMG-signals are considered as linear stationary that doesn't allow to characterize their time-frequency structure and objectively estimate the dynamic processes occurring in muscles.
To increase information content and expand of modern medicine possibilities it should be developed the automated complexes for processing and analysis of the biomedical signals, adequately reflecting processes in physiological systems. So for the analysis of a non-stationary EMG-signal by the nature it is offered to use the method of time-frequency transformation.
For the purpose of the analysis of time-frequency representation efficiency in objective qualitative and quantitative estimation of human neuromuscular system functional condition this work is given researches normal and pathologic global electromyogram, processed by the method of time-frequency transformation.
The qualitative analysis of structure of non-stationary EMG-signal by the nature (including the temporary location of its spectral components) and dynamics of its parameters in the process of muscle contraction performes on the basis of spectrogram. Spectrogram realizes the graphic visualization of amplitude, frequency and time components of biomedical signal.
The quantitative estimation of global electromyogram provides the calculation of average amplitude and parame-ters of time-frequency representation (median frequency, effective bandwidth of frequencies). Comparative analy-sis of the calculated parameters in normal and pathological conditions revealed a number of laws: indicators of median frequency and effective bandwidth of frequencies in pathology exceed the corresponding values in normal. This is due to infringement of coordinated work of muscle motor units, processes of desynchronization and functional reorganization of their activity at pathology.
On the basis of the established laws criteria for an estimation of human neuromuscular system functional con-dition are offered: the ratio of average amplitude to median frequency and (or) the ratio of average amplitude to effective bandwidth frequency. The given criteria allow taking into account the key indicators of bioelectric signal (amplitude and frequency) and thus carry out fast and effective express diagnostics of neuromuscular system functional condition with use of automated complexes of the EMG-signal time-frequency processing.