Extraction, Processing and Analysis of Surface Electromyogram Signal and Detection of Muscle Fatigue Using Machine Learning Methods

Muscle fatigue is a condition where a muscle or group of muscles lose their ability to contract and generate force. This can happen due to a variety of factors, including prolonged physical activity, lack of oxygen, and depletion of energy stores in the muscle. The raw sEMG signal is extracted by means of gel electrode attached to biceps of right arm. The preprocessing method used in the work involves different order of filters to process the raw signal. Further, the filtered signal is also amplified using instrumentation amplifier. The designed hardware extracts the signal at a frequency range between 56 Hz and 170 Hz. Six statistical features are extracted from the filtered signal in the time domain. The extracted features are given to various trained machine learning models using different algorithms such as Random Forest (RF), Support Vector Machine (SVM) and Logistic Regression (LR). The highest accuracy of about 87.5 % is achieved using random forest algorithm with the precision of 90%. The results that are obtained proves that machine learning methods can be used effectively to detect muscle fatigue from sEMG signals. The proposed method shows the propitious results in terms of accuracy and decisiveness. It can be used in areas such as sports training, rehabilitation, and ergonomics. This complete circuit is easy to produce and implement which could be used in the development of wearable and portable devices.