• JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
 
  Bookmark and Share
 
 
Doctoral Thesis
DOI
10.11606/T.76.2014.tde-31012014-095623
Document
Author
Full name
Osvaldo Vargas Jaques
E-mail
Institute/School/College
Knowledge Area
Date of Defense
Published
São Carlos, 2014
Supervisor
Committee
Costa, Luciano da Fontoura (President)
Cruvinel, Paulo Estevão
Kinouchi Filho, Osame
Liang, Zhao
Pereira, Cecilia Hedin
Title in Portuguese
Evolução do relacionamento entre dinâmica e topologia em redes neuronais: uma abordagem computacional
Keywords in Portuguese
Morfologia
Redes complexas
Redes neuronais
Relacionamento entre dinâmica e topologia
Abstract in Portuguese
Esta tese aborda o interrelacionamento entre morfologia, topologia e dinâmica de ativação em redes neuronais morfologicamente realistas, construídas com neurônios da base pública Neuromorpho. Foi desenvolvido um arcabouço computacional capaz de simular a dinâmica de ativação neuronal (através do modelo integra-e-dispara) ao longo do desenvolvimento da conexão das redes tridimensionais respectivas. Tal arcabouço permitiu investigar como aspectos da dinâmica de ativação variam ao longo das épocas de desenvolvimento das redes, incluindo antes, durante e depois da percolação. Em particular, calcularam-se correlações de Pearson entre várias medidas dinâmicas e topológicas ao longo das épocas de evolução, de forma a se quantificar de maneira objetiva os possíveis relacionamentos entre a ativação neuronal e a topologia das redes. Foram considerados três tipos de neurônios piramidais: occipitais e pré-frontais de humanos e células piramidais de macado (macaco Rhesus). Os dois primeiros tipos foram verificados (através de histogramas de médias e análise por componentes principais) possuírem características morfológicas semelhantes, enquanto o grupo de células piramidais do macaco apresentaram substancial diferenciação. Vários resultados foram obtidos respectivamente às correlações entre medidas dinâmicas e topológicas. Em particular, verificou- se que os graus médios de entrada e saída das redes estão significativamente correlacionados com as taxas médias de ativação, convergindo rapidamente após a formação do componente fraco. A correlação do grau de entrada mostrou-se mais elevada do que a obtida para o grau de saída. Além disto, a correlação entre o grau de entrada e a taxa média de disparos tendeu a diminuir ao longo das épocas finais das simulações. Verificou-se também como os perfis de evolução de várias correlações entre dinâmica e topologia implicam na diferenciação dos tipos neuronais considerados.
Title in English
Evolution of the relationship between dynamics and topology a computational approach
Keywords in English
Complex networks
Morphology
Neuronal networks
Relashionship between dynamics and topology
Abstract in English
This thesis addresses the interrelationships between morphology, topology and activation dynamics in morphologically realistic neuronal networks, derived from the public data base Neuromorpho. A computational framework has been developed that is capable of simulating the dynamics of neuronal activation (via the integrate-and-fire dynamics) during the development of the network connection in three-dimensional spaces. This framework allowed to investigate how aspects of the activation dynamics vary over the epochs of network development, including before, during and after the critical event of percolation. In particular, we calculated Pearson correlation coefficients between various topological and dynamical measurements throughout the epochs of evolution, in order to quantify in an objective way how the relationships between neuronal activation and network topology changed along the development of the connectivity. We considered three types of neurons: occipital and prefrontal pyramidal cells of human and diverse pyramidal cells of monkey individuals (monkey Rhesus). The first two types were found (via histograms and principal component analysis) to have mostly similar morphological characteristics, while the group of pyramidal cells from apes showed substantial differentiation. Several results were obtained respectively to the correlations between measurements of dynamics and morphology along the epochs of development. In particular, it was found that the input and output average degrees of the network are significantly correlated with the average rates of activation. After a period of large variation that precedes the formation of the weakly connected component, these correlation values converge rapidly to a regime of smooth decrease which suggests saturation of the activation in the network. The correlation implied by the indegree proved to be clearly higher than that obtained for the outdegree. It was also investigated how the profiles of the various correlations along the development epochs implied in the differentiation between the neuronal types considered.
 
WARNING - Viewing this document is conditioned on your acceptance of the following terms of use:
This document is only for private use for research and teaching activities. Reproduction for commercial use is forbidden. This rights cover the whole data about this document as well as its contents. Any uses or copies of this document in whole or in part must include the author's name.
Publishing Date
2014-02-04
 
WARNING: Learn what derived works are clicking here.
All rights of the thesis/dissertation are from the authors
Centro de Informática de São Carlos
Digital Library of Theses and Dissertations of USP. Copyright © 2001-2021. All rights reserved.