Computational modeling for gratings stimulated gamma oscillations in a large-scale cortical neuronal network

Biological experiments have shown that the oscillatory properties of gamma oscillation (30–80 Hz) in the primary visual cortex, which may be related to visual information transmission and integration, can be modulated by visual stimuli. However, the visual stimulus-modulation mechanisms of gamma oscillation are not fully unveiled. In this paper, a comprehensive computational model incorporating global drifting sinusoidal grating and a large-scale network of primary visual cortex is constructed to investigate this issue. The model reproduces the stimulus-dependent phenomena of gamma oscillation in a series of biological experiments, including the dependence phenomena of the spectral power and peak frequency of gamma oscillation on grating's orientation, contrast, and temporal frequency, respectively. In addition, the analysis of spike clusters and input current fluctuations found that the neuronal properties and particular function-specific connectivity patterns play important roles in modulating gamma oscillation in the cortex network. Specifically, the orientation selectivity of neurons determines the spectral power, whereas the push–pull scheme of synapses in layer 4 prevents the peak frequency from changing; the difference between excitatory and inhibitory neurons makes the peak frequency proportional to the grating's temporal frequency and causes the power turning point to vary with grating contrast. The study provides a comprehensive computational model and approach for deeply understanding the gamma oscillatory phenomena in the primary visual cortex. It can help to understand the dynamic mechanisms of visual stimulus-evoked gamma oscillations and help predict and guide relevant biological experiments; moreover, it may lay the foundation for further research on the cognitive function of related oscillations.