小世界神经网络中信号传输的建模研究

项目名称： 小世界神经网络中信号传输的建模研究

项目编号： No.60871094

项目类型： 面上项目

立项/批准年度： 2009

项目学科： 生物科学

项目作者： 李春光

作者单位： 浙江大学

项目金额： 30万元

中文摘要： 信号在大脑中的传输对于认知和计算等功能是必不可少的，但神经系统如何保证信号准确可靠地在脑中传输仍然不清楚。人们提出了信号传输主要的两种可能传输方式：同步脉冲(synfire)传输和放电速率(firing-rate)传输。对于这两种传输机制，在文献中的研究主要是基于规则的前馈网络或随机网络进行的，然而最近研究发现生物神经元网络既不是完全规则的也不是完全随机的，而是具有小世界性质的。在本项目中，我们系统地研究了小世界拓扑如何影响神经脉冲同步，进而如何影响神经信息的传输，研究了不同的激励突触和抑制突触比率下的信号传输，重点研究了不可靠突触对信号传输的影响。我们发现了小世界特性和不可靠突触的一些特殊作用，比如，在某些情况下神经突触的不可靠性反而对网络的同步和信息的传输更有利。在本项目的支持下，我们还对小世界神经元网络中对渐变模式的联想记忆、仿视觉神经系统的显著性检测和图像分割，以及神经元模型的电子电路实现等进行了研究。主要研究工作都属于神经科学和信息科学的交叉融合。

中文关键词： 信息传输；小世界网络；不可靠性；噪声；同步

英文摘要： The transmission of distributed neural activity within the brain is essential for high-level cognitive functions and computations, but it is not clear how the brain's architecture supports such signal propagation in a reliable and accurate manner. Recently, two alternative propagation modes have been discussed in the literature, namely the synfire chain and the firing-rate propagation. Investigations of both modes of signal transmission are either based on regular feed-forward or random networks. On the other hand, many biological networks including cortical systems are neither fully regular nor fully random. Rather, these networks have the so-called "small-world" property. In this project, we systematically investigated how the small-world topology affect the synchronization of neuronal networks and affect the signal propagation in the networks. We investigated how the ratio of excitatory vs inhibitory synapses affects the signal propagation. We investigated how the synaptic unreliability affects the signal propagation in the neuroanl networks. We found some interesting roles of the small-world topology and the synaptic unreliablity, for example, in some cases, the synaptic unreliability plays constructive roles on the synchronization and signal propagation. Supported partially by this foundation, we studied also the memorizing of morph patterns in small-world neuronal networks, the visual saliency detection algorithm, the bio-inspired image segmentation algorithm, as well as the electronic circuit design of the neuron models. All of these works are interdisciplinary researches between neuroscience and information science.

英文关键词： signal propagation; small-world network; unreliability; noise; synchronization