Abstract :
Parceled out into overlapping waveform pieces that mimic the mother wavelet in the time domain. However, if the goal of neural waveform analysis is to identify specific neural events or components, it is advantageous to choose wavelets that match the shapes of the signals of interest. The wavelet that most closely resembles the waveshape of a neural signal is the best scale-independent analyzing function for that signal in a matched filtering sense. Wavelets that are poor shape matches to a waveform or component will tend to produce more distributed energy patterns, complicating the detection, classification, and estimation of the signal of interest.
