An Approach to Capturing Neuron Morphological Diversity

This chapter 9 analyzes an algorithm for repairing computer-based neuronal reconstructions. It deals with a new probability density cloud approach for characterizing and repairing neuron dendritic and axonal morphologies. The chapter uses a layer 2 pyramidal cell to validate the dendritic repair process. In the repair algorithm, it assumes that the dendrites of neurons have a statistical symmetry in their morphologies, although they do not necessarily seem geometrically symmetric. The chapter also shows that the algorithm depends on the quality as well as the quantity of the population of cells.... This book offers an introduction to current methods in computational modeling in neuroscience, and describes realistic modeling methods at levels of complexity ranging from molecular interactions to large neural networks. A “how to” book rather than an analytical account, it focuses on the presentation of methodological approaches, including the selection of the appropriate method and its potential pitfalls. The book is intended for experimental neuroscientists and graduate students who have little formal training in mathematical methods, but will also be useful for scientists with theoretical backgrounds who want to start using data-driven modeling methods. The mathematics needed are kept to an introductory level; the first chapter explains the mathematical methods the reader needs to master to understand the rest of the book. The chapters are written by scientists who have successfully integrated data-driven modeling with experimental work, so all of the material is accessible to experimentalists. The chapters offer comprehensive coverage with little overlap and extensive cross-references, moving from basic building blocks to more complex applications.