Inferring neuronal network functional connectivity with directed information

Abstract

A major challenge in neuroscience is to develop effective tools that infer the circuit connectivity from large-scale recordings of neuronal activity patterns. In this study, context tree maximizing (CTM) was used to estimate directed information (DI), which measures causal influences among neural spike trains in order to infer putative synaptic connections. In contrast to existing methods, the method presented here is data driven and can readily identify both linear and nonlinear relations between neurons. This CTM-DI method reliably identified circuit structures underlying simulations of realistic conductance-based networks. It also inferred circuit properties from voltage-sensitive dye recordings of the buccal ganglion of This method can be applied to other large-scale recordings as well. It offers a systematic tool to map network connectivity and to track changes in network structure such as synaptic strengths as well as the degrees of connectivity of individual neurons, which in turn could provide insights into how modifications produced by learning are distributed in a neural network. This study brings together the techniques of voltage-sensitive dye recording and information theory to infer the functional connectome of the feeding central pattern generating network of In contrast to current statistical approaches, the inference method developed in this study is data driven and validated by conductance-based model circuits, can distinguish excitatory and inhibitory connections, is robust against synaptic plasticity, and is capable of detecting network structures that mediate motor patterns.

Reconstructing neuronal circuitry from parallel spike trains

“…The second approach, which we take here, is to use all of the data to carry out mesoscopic neuroanatomy, that is, to reveal the fine neuronal circuitry in which neural circuit computation is carried out. From these high channel count recordings, one should be able to estimate neuronal connectivity by quantifying the degree to which firing from a given neuron is influenced by the firing of neurons from which the index neuron is receiving input 14–30 . For this purpose, we develop an analytical tool that estimates neuronal connectivity in measurement units of postsynaptic potentials (PSPs).…”

Inferring neural information flow from spiking data

“…On the other hand, time-delayed versions of the CTW-based estimator were elaborated in [56] , [64] to infer task-driven directional interactions between the thalamus and the somatosensory area 1 (S1) in monkeys performing a tactile detection task [56] , and across cortical somatosensory, premotor and motor areas in monkeys performing a tactile dissemination task [64] . Finally, an extension of the CTW algorithm for non-necessarily finite-order Markov processes [65] was used to estimate the DI rate between neural spike trains from the buccal ganglion of Aplysia [66] .…”

Cadeias estocásticas de memória ilimitada com aplicação na neurociência

“…A fim de estudar a informação que flui de uma sequência de disparos neuronal a outra, usamos uma modelagem via cadeias estocásticas de memória ilimitada. Cai et al (2017) utiliza a informação dirigida -uma medida baseada em entropia -para medir as influências causais entre sequências de disparos neuronais a fim de inferir a conexão sináptica. Esta ferramenta pode ser utilizada para detectar conexões funcionalmente relevantes a partir de uma métrica baseada na informação fornecida pela gravação de disparos neuronais.…”

“…Enquanto a atividade neuronal pode ser observada diretamente, as interações entre as estruturas neuronais podem somente ser inferidas a partir dos dados. Cai et al (2017) propõem estimar a informação dirigida em pequenas redes neurais a partir de árvores de contexto maximizadas (ver, O'Neill et al (2012)) com base no estimador KT das probabilidades condicionais (ver, Krichevsky e Trofimov (1981)). Neste trabalho, entretanto, realizamos a estimação da informação dirigida a partir de estimadores empíricos.…”

“…A informação dirigida é uma medida baseada em entropia que tem muita semelhança com a informação mútua sendo que uma das diferenças é levar em conta uma parte casual de X ao longo da história da sequência Y em análise (CAI et al, 2017). Um interesse natural é estudar como o estimador empírico da informação dirigida se concentra em torno da sua taxa, já que essa é uma medida comumente usada na prática por ser limitada pela máxima entropia.…”