Detecting Evoked Responses under Stimulation Driven by Maximum-Length Sequences

Abstract

Objective Response Detection (ORD) Techniques applied to Evoked Potentials depend on the effective Number of Degrees of Freedom (NDOF) of spontaneous EEG. Previous works proposed a method for estimating the NDOF of the Evoked Potential Detector (EPD) probability distribution under the null hypothesis of no response, based on the autocorrelation function (ACF) of the EEG, modeled as wide- or narrow-band colored noise. Considering narrow-band EEG, the use of Beta distribution based on NDOF estimates via ACF is able to reflect EPD distribution provided that the stimulation rate is non-uniform. The present work assesses an approach for randomizing the inter-stimulus interval using Maximum-Length Sequences (MLS) based on Monte Carlo simulation for several combinations of EEG bandwidths, central frequencies, numbers of epochs and number of samples per epoch. The results suggest such approach as a valid algorithm to be implemented in the development of specific devices to produce random inter-stimuli intervals.