The application of the Hilbert spectrum to the analysis of electromyographic signals

摘要: This paper investigates the application of Hilbert spectrum (HS), which is a recent tool for analysis nonlinear and nonstationary time-series, to study electromyographic (EMG) signals. The HS allows visualization energy signals through joint time-frequency representation. In this work we illustrate use in two distinct applications. first feature extraction from EMG Our results showed that instantaneous mean frequency (IMNF) estimated relevant clinical practice. We found median IMNF reduces when force level muscle contraction increases. second investigated detection motor unit action potentials (MUAPs). MUAPs basic step decomposition tools, provide information about neuromuscular system morphology firing time MUAPs. compared, visually, how MUAP activity perceived on with visualizations provided by some traditional (e.g. scalogram, spectrogram, Wigner-Ville) distributions. Furthermore, an alternative HS, MUAPs, proposed compared similar approach based continuous wavelet transform (CWT). both technique CWT allowed clear distributions, whereas obtained were most difficult interpret as they extremely affected spurious activity.