Identifying significant frequencies in surface EMG signals for localization of neuromuscular activity

Presents two methods for determining significant frequencies in surface EMG recordings. In the first method, time-frequency distributions (TFDs) are analyzed to determine the ten most powerful frequencies (over time). The TFD frequencies are further analyzed using a “cyclostationary” approach. In the second method, AR modeling is used to evaluate how significant spectral components in the surface EMG signals change over time. These methods are important for identifying significant spectral components in multichannel surface EMG recording so that these excitations can be localized within the muscle