Skip to Main content Skip to Navigation
Theses

EXPLORATION PAR DES INTERFACES HYBRIDES DU CODE NEURONAL ET DES MÉCANISMES DE RÉGULATION DE L'INFORMATION SENSORIELLE DANS LE SYSTÈME VISUEL

Abstract : Determining the neural code in the thalamus and cerebral cortex, especially in the primary visual area, is hindered by the complexity of the neural network which is based on an astonishing diversity of neuronal types exhibiting varying properties at the morphological, biochemical, electrical and synaptic levels. This diversity is amplified by the numerous functional properties of each neuron which are reflecting the highly recurrent synaptic connections in cortical circuits as well as the corticothalamocortical loop. In other words, the response specificity of each neuron is affected through thousands of excitatory and inhibitory synapses, by the overall computation performed in the thalamocortical network. In the first part, we developed a model of contextual synaptic bombardment reproducing the dynamics of thousands of excitatory and inhibitory synapses converging to a single cortical neuron. A major advantage of this model is the possibility to control the amount of synchronization among the afferent synapses contacting the cortical neuron. We show in the visual cortex of cats that the amount of synaptic synchronization is related to the sub-threshold neuronal activity correlation level. Classically used artificial stimulations such as drifting gratings led the cerebral cortex into a redundant and correlated state while natural stimulations produced a richer neural code with less correlations. These results indicate that the sub-threshold neuronal activity correlation level is an indicator of the functional state in which the cerebral cortex is engaged. In the second part, we further investigated the neural code by extending our study to the thalamus, the major gateway for the flow of sensory information from the periphery to the cerebral cortex. The thalamus receives a strong corticothalamic feedback which results from the overall computation performed by the cortical areas. In order to study the impact of the corticothalamic feedack, we modeled a retinothalamocortical pathway mixing artificial and biological neurons recorded in the slice and we mimicked in these neurons a synaptic bombardment of cortical origin through the injection of mixed excitatory and inhibitory stochastic inputs in dynamic-clamp. This approach allowed us to control independently every thalamic neurons involved in the artificial pathway. We show that the sensory information transfer from the retina to the primary visual cortex is regulated by both a stochastic facilitation process across the population and the classical gain control a the cellular level. The stochastic facilitation process which could not be seen at the single-cell level is governed by the level of inter-neuronal correlation of the neuronal activity in the thalamus. Unlike conventional views, -a highly decorrelated neuronal activity- optimizes the sensory information transfer from the retina to the cortex by promoting the synchronization of synaptic inputs. We suggest that a cortically-induced decorrelation could increase the transfer efficiency for specific cell assemblies in the thalamus, constituting an attentional mechanism at the level of the thalamocortical circuits. At the same time, we developed a method to extract synaptic conductance fluctuations from single-trial intracellular recordings. We expect this method will help refining our understanding of the synaptic contexts in which the neurons are immersed with potential benefits on the development of new synaptic bombardment models. To conclude, our work confirms the hypothesis of a neural code based on synaptic synchronization governed by the level of correlation of the neuronal activity. Our results are consistent with numerous studies on attentional processes and suggest that active correlation and decorrelation mechanisms as well as oscillatory activities may regulate the information transfer between sensory organs and cortical areas.
Document type :
Theses
Complete list of metadatas

https://tel.archives-ouvertes.fr/tel-00714145
Contributor : Sébastien Béhuret <>
Submitted on : Tuesday, July 3, 2012 - 2:59:42 PM
Last modification on : Wednesday, December 9, 2020 - 3:14:36 PM
Long-term archiving on: : Thursday, October 4, 2012 - 3:00:10 AM

Identifiers

  • HAL Id : tel-00714145, version 1

Citation

Sébastien Béhuret. EXPLORATION PAR DES INTERFACES HYBRIDES DU CODE NEURONAL ET DES MÉCANISMES DE RÉGULATION DE L'INFORMATION SENSORIELLE DANS LE SYSTÈME VISUEL. Neurosciences. Université Pierre et Marie Curie - Paris VI, 2012. Français. ⟨tel-00714145⟩

Share

Metrics

Record views

407

Files downloads

602