Fatigue Assessment of Bicep Brachii Muscle Using Surface EMG Signals Obtained from Isometric Contraction

Abstract

In this study, Surface EMG signals are used to analyze the progression of muscular fatigue with time by estimating the change in myoelectric properties when right bicep brachii muscle is subjected to constant force isometric contraction. Muscular fatigue most frequently occurs due to powerful utilization of a group of muscles which can lead to decline in performance or sometimes to injury and can go undetected at early stage. In this proposed method, Discrete Wavelet Transform is used to decompose the EMG signals using Daubechies type 7 wavelet with three level of decomposition. For each detailed and approximate component temporal features like Root Mean Square, and Spectral features like Mean frequency, Median frequency and Energy are evaluated. Results show that mean frequency values perform significantly better in estimating the level of muscular fatigue with time. Furthermore, using Support Vector Machine classifier, the subjects were classified into muscular and non-muscular groups and second level detailed component shows high class separability in feature space.