中文摘要：

对于培养的神经元网络而言，学习是外界刺激与网络响应之间联系建立和调控的过程．为构建合适的神经元网络学习模型，采用闭环低频（1Hz）成对电极的电刺激模拟认知任务，在多通道微电极阵列系统中对培养的海马神经元网络进行训练，使其发生网络层次上的学习行为．经过训练后，神经元网络在刺激后20-80ms内的早期突触后响应明显增加，响应／刺激比（在闭环训练中，电极上任一阶段连续10次刺激的早期突触后响应的个数／10）增大，响应时延减小，并且响应具有选择性，即表明，神经元网络与外界刺激之间己建立可调控的联系，该可调控联系是通过网络的响应来表现的，建立神经元网络与外界刺激之间的可调控联系即网络层次的学习．

英文摘要：

Learning a new behavior is an exploration process that involves the formation and the modulation of association set between stimuli and neuronal networks＇ responses. To construct one appropriate learning model at networking level, on multi-electrode array, the closed-loop low frequency （1 Hz） pair stimulation was used to train the cultured hippocampal neuronal network and induce the learning behavior at networking level. Once the hippocampal neuronal network was trained successfully, the early postsynaptic responses were increased evidently, the response/stimulus ratio was increased and the response time decreased, furthermore, selective. These results indicate that the formation and the modulation of association between stimuli and neuronal networks＇ responses have been set, which means the learning behavior at networking level has occurred.