ABSTRACT

This chapter discusses the incorporating morphological information into account, trying to circumvent as many as possible of the above caveats and allowing for 2D and 3D characterization of single neurons and networks. It describes a number of concepts and techniques aimed at quantifying the morphometric characterization of neurons in order to lead towards the modeling and simulation of morphologically more realistic neural networks. Ramon-Moliner specifically cat retinal ganglion cells, by using the aforementioned shape features and two alternative clustering approaches are also presented, which is followed by the description of an approach to the synthesis, modeling and simulation of morphologically more realistic neural structures. As far as the former issue is concerned, it should be observed that the algorithmic approach adopted in the present work allows the characterization of a number of important shape features in neural cells, including complexity, area of influence, spatial coverage, and dendrograms.