In this thesis, we explore how different neuron populations interact with each other, and simulate the dynamics of the neurons with Continuous Attractor Neural Networks (CANNs). First we generalize the single-module CANNs to a bimodular structure, simulating two sensory modalities in the neural network. Second, we explore the dynamics of bimodular CANNs when applied with different kinds of stimuli and under various neuron couplings. We begin with applying two static stimuli on two modules respectively in the network model, and adjust the inter-modular couplings to vary the influence between the two neural modalities. We found that there is competition between the external input and the inter-modular couplings within each neural modality. We further study the dynamics of the netw...[ Read more ]

In this thesis, we explore how different neuron populations interact with each other, and simulate the dynamics of the neurons with Continuous Attractor Neural Networks (CANNs). First we generalize the single-module CANNs to a bimodular structure, simulating two sensory modalities in the neural network. Second, we explore the dynamics of bimodular CANNs when applied with different kinds of stimuli and under various neuron couplings. We begin with applying two static stimuli on two modules respectively in the network model, and adjust the inter-modular couplings to vary the influence between the two neural modalities. We found that there is competition between the external input and the inter-modular couplings within each neural modality. We further study the dynamics of the network with one static stimulus and one moving stimulus. The module with moving stimulus shows a wide range of dynamics in its tracking behavior depending on inter-modular coupling. We fit the bimodular CANN model to a sensory illusion experiment, simulating interactions between two different sensory modalities, illustrating the probable mechanism behind the biological phenomenon.

We are also interested in the couplings between cells in the retina. Our collaborator, Prof. C. K. Chan’s group from Institute of Physics, Academia Sinica did experiments on the bullfrog retina. They found that when the bullfrog retina is shown a moving bar driven by the Hidden Markov Model (HMM), the responses of the ganglion cells contain information about the future inputs, meaning the retina is not merely a station receiving signals and transferring information to the visual cortex, but is also able to make predictions about future inputs. This predictive effect is due to the inertia in the HMM information. If the moving bar positions do not contain momentum or inertia information, as in the Ornstein–Uhlenbeck process (OU process), the retina will lose the ability to make predictions. In our simulations we consider two neural populations in the neural network possibly corresponding to ganglion cells and amacrine cells. Our simulation results show that our neural network model fits with the experiments very well.