当前位置:
X-MOL 学术
›
Phys. Rev. E
›
论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
Information processing in tree networks of excitable elements
Physical Review E ( IF 2.4 ) Pub Date : 2021-01-22 , DOI: 10.1103/physreve.103.012308 Ali Khaledi-Nasab 1 , Kanishk Chauhan 2 , Peter A Tass 1 , Alexander B Neiman 2, 3
Physical Review E ( IF 2.4 ) Pub Date : 2021-01-22 , DOI: 10.1103/physreve.103.012308 Ali Khaledi-Nasab 1 , Kanishk Chauhan 2 , Peter A Tass 1 , Alexander B Neiman 2, 3
Affiliation
We study the collective response of small random tree networks of diffusively coupled excitable elements to stimuli applied to leaf nodes. Such networks model the morphology of certain sensory neurons that possess branched myelinated dendrites with excitable nodes of Ranvier at every branch point and at leaf nodes. Leaf nodes receive random inputs along with a stimulus and initiate action potentials that propagate through the tree. We quantify the collective response registered at the central node using mutual information. We show that in the strong-coupling limit, the statistics of the number of nodes and leaves determines the mutual information. At the same time, the collective response is insensitive to particular node connectivity and distribution of stimulus over leaf nodes. However, for intermediate coupling, the mutual information may strongly depend on the stimulus distribution among leaf nodes. We identify a mechanism behind the competition of leaf nodes that leads to nonmonotonous dependence of mutual information on coupling strength. We show that a localized stimulus given to a tree branch can be occluded by the background firing of unstimulated branches, thus suppressing mutual information. Nonetheless, the mutual information can be enhanced by a proper stimulus localization and tuning of coupling strength.
中文翻译:
可兴奋元素树网络中的信息处理
我们研究了扩散耦合的可激发元素的小型随机树网络对应用于叶节点的刺激的集体反应。这种网络模拟某些感觉神经元的形态,这些神经元具有分支有髓树突,在每个分支点和叶节点具有可兴奋的 Ranvier 节点。叶节点接收随机输入和刺激,并启动通过树传播的动作电位。我们使用互信息量化在中央节点注册的集体响应。我们表明,在强耦合极限下,节点数和叶子数的统计决定了互信息。同时,集体响应对特定节点的连接性和叶节点上的刺激分布不敏感。然而,对于中间耦合,互信息可能强烈依赖于叶节点之间的刺激分布。我们确定了叶节点竞争背后的机制,该机制导致互信息对耦合强度的非单调依赖。我们表明,给予树枝的局部刺激可以被未受刺激的树枝的背景发射所遮挡,从而抑制互信息。尽管如此,互信息可以通过适当的刺激定位和耦合强度的调整来增强。从而抑制互信息。尽管如此,互信息可以通过适当的刺激定位和耦合强度的调整来增强。从而抑制互信息。尽管如此,互信息可以通过适当的刺激定位和耦合强度的调整来增强。
更新日期:2021-01-22
中文翻译:
可兴奋元素树网络中的信息处理
我们研究了扩散耦合的可激发元素的小型随机树网络对应用于叶节点的刺激的集体反应。这种网络模拟某些感觉神经元的形态,这些神经元具有分支有髓树突,在每个分支点和叶节点具有可兴奋的 Ranvier 节点。叶节点接收随机输入和刺激,并启动通过树传播的动作电位。我们使用互信息量化在中央节点注册的集体响应。我们表明,在强耦合极限下,节点数和叶子数的统计决定了互信息。同时,集体响应对特定节点的连接性和叶节点上的刺激分布不敏感。然而,对于中间耦合,互信息可能强烈依赖于叶节点之间的刺激分布。我们确定了叶节点竞争背后的机制,该机制导致互信息对耦合强度的非单调依赖。我们表明,给予树枝的局部刺激可以被未受刺激的树枝的背景发射所遮挡,从而抑制互信息。尽管如此,互信息可以通过适当的刺激定位和耦合强度的调整来增强。从而抑制互信息。尽管如此,互信息可以通过适当的刺激定位和耦合强度的调整来增强。从而抑制互信息。尽管如此,互信息可以通过适当的刺激定位和耦合强度的调整来增强。