标题： 异质神经回路中通过递归抑制的去相关作用 显示英文标题

标题： Decorrelation by recurrent inhibition in heterogeneous neural circuits

摘要： 皮层神经元的活动是相关的，这种相关性影响大脑处理信息的方式。 什么神经回路机制导致了这种相关性？ 共享的突触输入对于同步神经元是必要的，但如果神经回路具有异质性，这可能并不足够。 我们通过分析一个由强而异质的相互作用特征化的网络模型来研究相关性的神经回路机制。 我们考虑强的兴奋性和抑制性连接：兴奋性输入驱动神经活动的波动，这些波动被抑制性反馈所抵消。 虽然强的兴奋性输入倾向于使神经元相关，但抑制性反馈强烈地减少了相关性。 关键的是，突触连接的异质性对于抑制相关性是必要的。 在这些条件下，我们发现相关性为正值，其大小为 K^{-1/2}，其中 K 是神经元的连接数。 这一结果与皮层的解剖和生理测量一致：皮层神经元接收大约 K \simeq 10^4 个连接，测得的相关性约为 10^{-2}。 显示英文摘要

摘要： The activity of cortical neurons is correlated, and this correlation affects how the brain processes information. What are the neural circuit mechanisms underlying correlations? Shared synaptic input is necessary to synchronize neurons, but it may not be sufficient if neural circuits are heterogeneous. We study the neural circuit mechanisms of correlations by analyzing a network model characterized by strong and heterogeneous interactions. We consider strong excitatory and inhibitory connections: the excitatory input drives the fluctuations of neural activity, which are counterbalanced by the inhibitory feedback. While the strong excitatory input tends to correlate neurons, the inhibitory feedback strongly reduces the correlations. Crucially, the heterogeneity of synaptic connections is necessary for the inhibition of correlations. Under these conditions, we found that correlations are positive and of magnitude K^{-1/2}, where K is the number of connections to a neuron. The result agrees with anatomical and physiological measurements in the cortex: a cortical neuron receives about K \simeq 10^4 connections, and the measured correlations are about 10^{-2}.