TIME DOMAIN HIGH-ORDER SPECTRAL ANALYSIS TO STUDY NEURAL PLASTICITY IN AN ASSOCIATIVE LEARNING TASK

In Neurosciences, efforts have been made to understand the related mechanisms of oscillatory neural activity through electrophysiological techniques, where time-frequency analysis has played a crucial role in unraveling the dynamics of neural activity associated with cognitive processes. However, in traditional approaches, such as power spectral analysis , the focus is solely on power distribution based on frequency, disregarding phase information, assuming a linear distribution, and ignoring the interaction of power with signal components that manifest with phase, a nonlinear phenomenon inherent in the central nervous system. In this sense, the importance of quantifying the processing of non-linear neural activity has encouraged the use of mathematical tools. The high-order spectral analysis in the time domain (TDHOSA) can provide useful information on possible nonlinearities and frequency coupling in systems and time series, beside time delay estimation and suppression of Gaussian noise with unknown spectral characteristics.