FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W606899741> ?p ?o ?g. }

• W606899741 endingPage "30" @default.
• W606899741 startingPage "27" @default.
• W606899741 abstract "yNeuroscience Laboratory, Department of Applied Mathematics, School of Optics, Universidad Complutense de Madrid;Avda. Arcos de Jalon s/n, 28037, Spain.zLaboratory of Nonlinear Systems, Swiss Federal Institute of Technology Lausanne;EPFL-IC-LANOS, CH-1015 Lausanne, Switzerland.Email: vmakarov@opt.ucm.es, oscar.defeo@epfl.ch, fivos.panetsos@opt.ucm.esAbstract—A novel method for the identification and mod-eling of neural networks using experimental spike trains isdiscussed. The method assumes a reference model of inter-connected deterministic integrate-and-fire neurons and fitthe parameters of the model to the observed experimentalspike trains. The identification provides the properties ofthe individual synapses and neurons, hence extracting thefunctional connectivity between neurons. The method isshown to be e ective when applied on simulated data.1. IntroductionThe qualitative and quantitative analysis of the spiking ac-tivity of individual neurons is a very valuable tool for thestudy of the dynamics and functional architecture of theneural networks in the Central Nervous System [1]. In par-ticular, deducing the functional connectivity of neural net-works from experimental data, usually restricted to spiketrains, is of crucial importance in neuroscience: for thecorrect interpretation of the electrophysiological activity ofthe involved neurons and networks; and, more important,for correctly relating the electrophysiological activity to thefunctional tasks accomplished by the network. Here, theterm functional stands for any observable, direct or indi-rect, interaction between neurons which alters their spiketimings.The measured activity of a neuron is not the result ofits solely intrinsic properties, but stems from the directand indirect influences of the other neurons of the net-work, leading to network behaviors far beyond the sim-ple combinations of those of the isolated neurons. On theother hand, the measured time instants of spike occurrences(point events) do not allow any direct insights about thesubthreshold and/or intrinsic membrane dynamics of theneurons. Nonetheless, spike trains can be used to identifythe functional characteristics and e ective architecture ofthe neural network they originated from, e.g [2, 3].The most common and standard methods for identifyingthe synaptic connections between neurons assume a sto-chastic nature of the spike trains, and the functional bondbetween two neurons is extracted from the statistical infer-ence of the discharges times, usually deducing it from theshapes of cross-correlograms [2, 3]. Though widely under-stood, this tool provides a very limited knowledge aboutthe functional properties of the neural networks, and itcannot distinguish direct from indirect connections amongneurons. Recently, more sophisticated statistical methods[4, 5, 6] have overtaken this problem. However, these meth-ods still fit into a stochastic framework and lack a com-pact description of the estimated interactions. Furthermore,since they assume a stochastic nature of the spike trains,they do not provide considerations about the dynamics ofthe involved neurons, nor about the nature of the intrinsicprocesses that are responsible for such behavior.In contrast to a purely statistic approach, a deterministicone can be considered, with the main advantage of pro-viding a mathematical model for inferring single neuronor neural network properties indirectly. In this direction,methods for extracting a dynamical system out of the inter-spike intervals have been recently proposed [7, 8], howeverthese methods assume neurons to be isolated; hence, theydo not provide insights about the neural network structureand its relationships with the observed dynamics.Here a new model based method for the identificationand modeling of whole neural networks from experimentalspike trains is proposed. A description of the method isgiven in Sec. 2, whilst in Sec. 3 numerical tests of it arepresented and then discussed in Sec. 4.2. Identification MethodThe identification method, despite being quite mathemati-cally convoluted, is rather transparent in its principle. Thereference model adopted in the identification process is anetwork of interconnected integrate-and-fire neurons. Allthe parameter values necessary to univocally define it, i.e.the connectivity matrix of the network, the synaptic timescales, and the intrinsic parameters of the neurons, are de-rived from the recorded spike trains by an optimizationprocedure which minimizes the di erence between the pre-dicted and measured timings of spike episodes.Precisely, given N spike series from as many neurons,the reference model is a network composed of N intercon-nected single-compartment leaky integrate-and-fire mod-els. The connections between neurons are represented by aN N matrix W whose elements w" @default.
• W606899741 created "2016-06-24" @default.
• W606899741 creator A5019394263 @default.
• W606899741 creator A5046235795 @default.
• W606899741 creator A5066505528 @default.
• W606899741 date "2004-01-01" @default.
• W606899741 modified "2023-10-02" @default.
• W606899741 title "Inferring Neural Connectivity and the Underlying Network Dynamics from Spike Train Recordings" @default.
• W606899741 cites W1586810339 @default.
• W606899741 cites W1987994242 @default.
• W606899741 cites W1988941949 @default.
• W606899741 cites W2091896507 @default.
• W606899741 cites W2093615513 @default.
• W606899741 cites W2094542824 @default.
• W606899741 cites W2129086996 @default.
• W606899741 cites W2132739442 @default.
• W606899741 cites W2164653071 @default.
• W606899741 cites W3148499377 @default.
• W606899741 hasPublicationYear "2004" @default.
• W606899741 type Work @default.
• W606899741 sameAs 606899741 @default.
• W606899741 citedByCount "0" @default.
• W606899741 crossrefType "journal-article" @default.
• W606899741 hasAuthorship W606899741A5019394263 @default.
• W606899741 hasAuthorship W606899741A5046235795 @default.
• W606899741 hasAuthorship W606899741A5066505528 @default.
• W606899741 hasConcept C115903868 @default.
• W606899741 hasConcept C118403218 @default.
• W606899741 hasConcept C118615104 @default.
• W606899741 hasConcept C119857082 @default.
• W606899741 hasConcept C123757187 @default.
• W606899741 hasConcept C154945302 @default.
• W606899741 hasConcept C15744967 @default.
• W606899741 hasConcept C169760540 @default.
• W606899741 hasConcept C185263204 @default.
• W606899741 hasConcept C2781390188 @default.
• W606899741 hasConcept C2909946758 @default.
• W606899741 hasConcept C33923547 @default.
• W606899741 hasConcept C41008148 @default.
• W606899741 hasConcept C50644808 @default.
• W606899741 hasConceptScore W606899741C115903868 @default.
• W606899741 hasConceptScore W606899741C118403218 @default.
• W606899741 hasConceptScore W606899741C118615104 @default.
• W606899741 hasConceptScore W606899741C119857082 @default.
• W606899741 hasConceptScore W606899741C123757187 @default.
• W606899741 hasConceptScore W606899741C154945302 @default.
• W606899741 hasConceptScore W606899741C15744967 @default.
• W606899741 hasConceptScore W606899741C169760540 @default.
• W606899741 hasConceptScore W606899741C185263204 @default.
• W606899741 hasConceptScore W606899741C2781390188 @default.
• W606899741 hasConceptScore W606899741C2909946758 @default.
• W606899741 hasConceptScore W606899741C33923547 @default.
• W606899741 hasConceptScore W606899741C41008148 @default.
• W606899741 hasConceptScore W606899741C50644808 @default.
• W606899741 hasLocation W6068997411 @default.
• W606899741 hasOpenAccess W606899741 @default.
• W606899741 hasPrimaryLocation W6068997411 @default.
• W606899741 hasRelatedWork W1679567753 @default.
• W606899741 hasRelatedWork W1896510368 @default.
• W606899741 hasRelatedWork W1900199311 @default.
• W606899741 hasRelatedWork W1980773744 @default.
• W606899741 hasRelatedWork W2014293473 @default.
• W606899741 hasRelatedWork W2039216706 @default.
• W606899741 hasRelatedWork W2042686653 @default.
• W606899741 hasRelatedWork W2056096402 @default.
• W606899741 hasRelatedWork W2062240204 @default.
• W606899741 hasRelatedWork W2063519073 @default.
• W606899741 hasRelatedWork W2072264193 @default.
• W606899741 hasRelatedWork W2105728001 @default.
• W606899741 hasRelatedWork W2112705928 @default.
• W606899741 hasRelatedWork W2118928905 @default.
• W606899741 hasRelatedWork W2131893090 @default.
• W606899741 hasRelatedWork W2147008284 @default.
• W606899741 hasRelatedWork W2147792971 @default.
• W606899741 hasRelatedWork W25580397 @default.
• W606899741 hasRelatedWork W2623691432 @default.
• W606899741 hasRelatedWork W1752608841 @default.
• W606899741 isParatext "false" @default.
• W606899741 isRetracted "false" @default.
• W606899741 magId "606899741" @default.
• W606899741 workType "article" @default.