各位老师，您好：

由实验室吴思老师邀请了上海交通大学特别研究员周栋焯博士来实验室做报告。报告信息如下，欢迎感兴趣的老师和同学来听。

报告时间：5月17日上午10:00

报告地点：英东楼422

报告题目：Neuronal dynamics from dendritic  integration to population responses

报告摘要：        In this talk, I will present  our modeling work about neuronal dynamics in comparison with experimental  results including (1) How a single neuron integrates excitatory and inhibitory  inputs from other neurons in the cortical network, and (2) How population  neurons emerge spatiotemporal activities in response to some simple external  stimulus. The first part discusses a simple arithmetic rule recently discovered  in experiments as $V^{Exp}_S≈V^{Exp}_E + V^{Exp}_I + kV^{Exp}_E V^{Exp}_I$ ,  where V^{Exp}_S, V^{Exp}_E and V^{Exp}_I are the amplitude of the summed somatic  potential (SSP), the excitatory postsynaptic potential (EPSP) and the inhibitory  postsynaptic potential (IPSP) measured at the time when the EPSP reaches its  peak value. In addition, k is defined as the shunting coefficient which only  depends on the spatial location of excitatory and inhibitory inputs on the  dendrite and is independent of the amplitude of EPSP and IPSP. We demonstrate  both theoretically and numerically that the above dendritic integration rule can  be explained by sub-threshold membrane potential dynamics as characterized in  the conductance-based integrate-and-fire (I&F) model. In order to account  for the spatial dependence of the shunting coefficient k, we propose a  dendritic-integration-rule-based integrate-and-fire (DIF) model. Our analytical  and numerical results show that this model is able to capture many experimental  observations. The second part discusses how we use a large-scale  conductance-based, integrate-and-fire neuronal network model in V1 to  investigate the spatiotemporal activities of population neurons in response to  some simple Garbor stimulus and compared our simulation results to the  experimental observations. Our model incorporates both isotropic local couplings  of AMPA and GABA, in addition, we model the lateral orientation-specific  long-range connections by a slow NMDA component. From the analysis of both  experimental observations and our numerical results, we demonstrate that the  time scale and the special scale of NMDA component play an essential role in  terms of special-temporal patterns and also the cortical correlation  structures.

报告人简介：

周栋焯，分别于2002年与2007年获北京大学数学科学学院学士学位与博士学位，博士期间主要研究方向是高分子流体的多尺度建模与计算，2007年至2009年在访问纽约大学库朗研究所期间，研究方向转为大脑神经科学中的数学建模和科学计算，并于2010年加入上海交通大学，任特别研究员。目前主要研究兴趣是：生物及物理中的应用数学问题，如高分子流体的多尺度建模与计算，神经元网络动力学的数学性质与方法的研究，生物与物理系统中的非线性与混沌现象，大尺度的视觉神经网络的建模与模拟的研究。

    此致

敬礼！

认知神经科学与学习国家重点实验室

2012年5月9日