Automated Extraction of Activity Features in Linear Envelopes of Locomotor Electromyographic Patterns

For the envelope of the electromyographic activity recorded during locomotion, a feature extraction scheme has been developed. It recognizes the time, duration, and amplitude of phases of activity. The method is based on the Tauberian approximation for modeling waveforms as a sum of identically shaped pulses with different time delays and amplitudes. Initial conditions on the pulse properties are set and an optimal solution is sought. Any multiple pulses that model one phase of activity are then combined into one equivalent pulse. The method is completely automated; thus, it operates without any user intervention.