Learning And Memory Model Of Vestibulo-ocular Reflex Circuit

With the development of brain science research,the cerebellum has attracted the attention of researchers due to its important role in motor learning and memory.Vestibulo-Ocular Reflex(VOR)is an eye-movement with self-adaptive adjustment caused by a combination of head rotation and retinal image slip,which is mainly controlled by VOR circuit.So far,researchers have proposed many learning and memory models of vestibular ocular reflex circuit.Although these models can partially reproduce the results of physiological experiments,it is still of great significance to study the mechanism of distributed plasticity site in VOR circuit and to establish a new model of learning and memory in VOR circuit.This thesis focuses on learning and memory models of VOR circuit and the main contents include:(1)There are multiple plastic synapses in the cerebellar VOR circuit.How these distributed synapses work together to learn and remember are still under study.Based on the analysis of the existing models,this thesis proposes a neuron activity-based learning and memory model in VOR circuit.In order to explore the contribution of different plasticity sites,the phase reversal experiments in normal and mutant mice were simulated.The simulation results show that the proposed model not only reproduces the physiological experiment data of normal mice well but also predicts learning and memory defects of mutant mice in comparison with the classical single plasticity site model.(2)The timing and amplitude control of the eye movement by VOR circuit represents the learning and memory ability of the circuit in time and space,respectively.In order to further study the learning and memory characteristics of VOR gain,this thesis proposes a learning and memory model based on spiking neural networks for VOR circuit.The simulation results verify that both gain-up stimulation and gain-down stimulation induce learning in the circuit.Meanwhile,this model reproduces the learning and memory effects of VOR circuit in space training.Compared with the classical model,the simulation results show that our model is well consistent with the physiological experiment data.In summary,this thesis studies classical models of VOR circuit and proposes learning and memory models of VOR circuit towards the phase-adaptation and gain-adaptation of eye movement.The proposed models reproduce the physiological experiment data of various training paradigms and validate the effect of distributed plasticity sites on learning and memory of VOR circuit.