E-Locus - Institutional Repository of the University of Crete

Home Search

Results - Details

Search command : Author="Ποϊράζη" And Author="Παναγιώτα"

Current Record: 9 of 31

[Add to Basket]

Identifier

000433840

Title

Intrinsic computations of a wide range of Boolean functions in apical dendritic branches of a two-layer abstract human pyramidal neuron model

Author

Παπαδόπουλος, Σωτήριος

Thesis advisor

Ποϊράζη, Παναγιώτα

Abstract

Pyramidal neurons form the substrate of elaborate, high-order cognitive functions of humans. Thus, our ability to understand how these neurons operate is crucial for deciphering brain functions, but it is constrained by technological barriers, especially in the microscale level. The use of modeling approaches has been proven very useful in studying neuronal mechanisms and elucidating the key parameters that govern them, when our experiments fail to do so. In this study, I have modeled a pyramidal neuron dendrite, as well as its soma, under the assumption that they their inputs can be described by binary values. The inputs are weighted and integrated resulting in binary outputs, enabling me to describe the input-output function of the dendrite or the neuron in terms of Boolean Algebra and truth tables. I have investigated how the computational capacity of a single dendrite -and hence the neuron- is affected by the characteristics of the dendritic input-output function. I show that a nonmonotone function provides access to the whole range of computations that can be assigned to a given number of inputs. Moreover, I reveal that the these computations are state-dependent. This analysis, therefore, provides insights into how human pyramidal neurons may be able to perform complicated calculations like non-linearly separable classification

Language

English

Subject

Abstract neuron models

Anti-coincidence detection

Boolean Algebra

Dendrites

Dendritic calcium-mediated action potential (dCaAP)

Dendritic computations

Dendritic spikes

Non-monotone activation function

Pyramidal neuron