Thesis (M.Sc.) - Cairo University - Faculty of Engineering - Department of Mathematics and Physics

Neuronal dendrites hold the largest portion of neuron spatial representation and bifurcation details of almost all neurons types. That{u2019}s why neurobiologists use dendrite morphologies as a classifier between different neuron types in different brain regions and layers. Based on the speculations of Ramón y Cajal about how dendrites take their spatial representation by optimizing the total wiring length and the conduction time within the dendritic tree, we generalized his hypothesis by adding the geometric boundaries that surround neuron cells into account; it significantly boosts the electrotonic matching of a synthetic dendritic tree with its real reconstruction. In this work, we added a new attribute that considers the change in the branch diameters (branching order) to the total number of attributes that control the growth of synthetic trees. These three attributes determine how the dendrites bifurcate. Also, we created two metrics that measure how much a synthetic dendritic tree and the real one are electrotonically and geometrically matched. The present hypothesis significantly boosts the electrotonic and geometric matching of a synthetic dendritic tree with its real reconstruction. Optimal values of the parameters are obtained, and, hence, the present technique is applicable as a quantitative classifier for different neuronal cells

Issued also as CD