Computational Study of Stimulus-Induced Synchrony in the Cat Retina
Synchronization of neuronal responses across two or more neurons is ubiquitous in the visual system. Including other species, it has been observed in the retina (Mastronarde, 1989), the lateral geniculate nucleus (Alonso, et al., 1996), and the visual cortex (Engel et al., 1991a) of the cat. Although a robust phenomenon, the role of synchronization is poorly understood. This dissertation investigates stimulus-induced synchrony in the cat retinal ganglion cells. For this purpose two modeling approaches are employed. First, a two-layered feedforward neural network model of a circuit that connects photoreceptor cells to X type ganglion cells in cat retina is presented. The first layer simulates the behavior of photoreceptor and bipolar cells, and its spatial response is modeled descriptively using a difference of gaussians function. The second layer simulates the behavior of ganglion cells and is modeled mechanistically using a noisy, leaky integrate and fire model. The model is constrained by the experimental data, and is used to study correlated spiking activity between the neighboring ganglion cells under uniform illumination. The model shows that common input only to the ganglion cells does not account for the experimentally observed correlations on the time scale of 2-10 ms between X type neighboring ganglion cells. However, a gap junction (electrical coupling) of conductance 0.02 mS/cm ^2 between the ganglion cells gives rise to the observed correlations between their spike trains. Second, the neural circuitry underlying the receptive fields of the retinal ganglion cells is neglected, and the ganglion cells are modeled by a more realistic conductance-based mathematical model. For Ornstein-Uhlenbeck noise as stimulus, the firing characteristics of a single model neuron are explored as the parameters of the stimulus are varied. Moreover, the responses of the uncoupled neurons to common Ornstein-Uhlenbeck noise with respect to the stimulus-induced phase synchronization are studied. The dependence of phase synchronization on the noise variance and frequency contents is explored. It is shown that optimal phase synchronization is achieved for a certain cut off frequency of the common noise source. For physiologically relevant stimuli, however, it is found that a gap junction coupling is required between the pair of neurons to realize tight synchrony.
School Location:USA - Ohio
Source Type:Master's Thesis
Keywords:retinal ganglion cells phase synchronization ornstein uhlenbeck noise stimulus induced synchrony bistable system hodgkin hurley
Date of Publication:01/01/2004