STRATEGY FOR REDUCING MUSCLE FATIGUE IN THE USE OF BIOELECTRIC PROSTHESIS BASED ON ELECTROMYOGRAPHIC ANALYSIS

Muscle fatigue is characterised by a decrease in the ability of muscles to perform required functions, which is accompanied by accumulation of metabolites, changes in ionic balance and electrical properties of muscle tissue. These physiological changes lead to marked transformations of electromyographic (EMG) signals, including a shift of their spectrum to lower frequencies. Such changes affect the quality of signal recognition and control accuracy of bioelectric prostheses, since most myoelectric control algorithms are based on the stability of spectral and amplitude characteristics of EMG signals. This paper investigates the influence of muscle fatigue on the characteristics of EMG signals and the accuracy of myoelectric control in bioelectric upper limb prostheses. The physiological mechanisms underlying changes in EMG signals during fatigue are analysed, including a slowing of the rate of conduction of action potentials along muscle fibres and changes in the recruitment of motor units. These factors result in decreased median frequency and increased amplitude of EMG signals, making it difficult to accurately classify and control them with a prosthesis. To address this problem, a method of muscle fatigue compensation based on adjusting the median frequency of the EMG signal is considered. The method is a spectral analysis of the current EMG signal to determine its median frequency and subsequent scaling of the time axis of the signal. This allows shifting the signal spectrum back to the higher frequencies characteristic of the fatigued state of the muscle and restoring the original spectral characteristics of the signal. Keywords: Bioelectric prosthesis, Electromyographic signals (EMG), myoelectric control, muscle fatigue, Adaptive classification algorithms, Biosignal processing