Modeling the Formation of Steady State Visual Evoked Potentials at Different Frequencies of Photostimulation

Abstract

The formation of steady-state visual evoked potentials on the basis of visual evoked potentials was analyzed. Using the regression model, it was shown that the accumulation of steady-state visual evoked potentials depends in part on the time position of visual evoked potentials, as suggested by different regression coefficients set for visual evoked potentials with different relative phase shifts. The description of steady-state visual evoked potentials using a linear regression model with iteration of visual evoked potentials was better at the photostimulation frequency of 14 Hz than at the frequency of 8 Hz. In the vast majority of cases, the number of groups of frequency peaks (chains of local maxima) detected on the spectra in the frequency–time domain was greater for signals induced at 8 Hz than for those induced at 14 Hz. This relationship was maintained both for steady-state visual evoked potentials recorded directly in test subjects and for signals artificially reconstituted from visual evoked potentials. For photostimulation at 14 Hz, the frequency components of the signal that increased and decreased their period were equally common, while at 8 Hz the number of chains of local maxima increased and their frequency decreased. It was shown that a group of subjects exists whose response to high-frequency photostimulation involves a decrease in the number of pronounced frequency components that contribute to the formation of the steady state evoked potential. At the same time, the modulation of frequency components in the native steady-state visual evoked potentials was predominantly characterized by a decrease in the signal frequency, which was not observed in the model signal reconstituted from a series of visual evoked potentials of the same subject.