Nonlinearity in the Light Processing of the Human Visual System

Abstract

Visual evoked potentials, i.e responses to visual stimulation as recorded using electroencephalography have indicated the existence of nonlinear behavior of the visual pathway. Nonlinearities and time delay in the visual system play an important role in understanding the complex nature of the visual system. This study investigated the nonlinear interactions and time delay in the visual pathway, using several types of stimulation paradigms. Multisine (i.e sum of multiple sinusoidal signals) and sine light stimulation were presented to healthy participants in order to elicit steady-state visual evoked responses. The recorded signals were analyzed using multi-spectral phase coherence, a novel cross-frequency phase coupling metric, in order to quantify the nonlinear interactions, form a brain map and estimate time delay. Chirp light stimulation (signals with linearly increasing frequencies) was used to elicit visual responses inside a specific frequency range. Time delay was estimated using Fractional Fourier Transformation due to its ability to handle chirps’ non-stationary properties. Brain maps indicated that multisine paradigms elicit more localized nonlinear interactions than chirp paradigms. All sinusoidal stimulation provided clusters of similar time delays. Bisine presented the most distinctive groups, fact that suggests that bisine is able to be used as a distinction measure. Trisine time delay showed the lowest variance, fact that shows more accurate estimation. Chirp time delay presented also small variances but the mean time delay found to be very frequency dependent. To conclude, this study showed that multisine paradigms are suitable to be used to elicit nonlinear responses but time delay may not be sufficient measure to fully describe the visual system.