Bispectral Analysis Of Filtered Impulse Processes With Applications To The Analysis Of Bioelectric Phenomena

This article presents two applications of hi-spectra for the analysis of bioelectric phenomena modeled as filtered impulse processes; the surface elect romyogram and spent aneous synaptic potentials. Estimation of the elementary waveform and the frequency of its occurrence is important for the analysis of such phenomena when the frequency of elementary waveform occurrence is high enough to cause heavy waveform contamination. Explicit expressions of the elementary waveform and the frequency of its occurrence in terms of the power spectrum and hi-spectrum are first presented in the case that the waveform occurrence times form a stationary Poisson process. Then the computer simulation demonstrates the applicability of the method for the analysis of bioelectric phenomena under consideration. The estimation of the elementary waveform may be utilized as a means of making noninvasive diagnoses of neuromuscular disorder and estimation of the frequency of elementary waveform occurrence may be applied to the analysis of transmitter release at bio-synapses.