Modeling neuron spatial distribution and morphology in the developing human cerebral cortex

Abstract Forests of synthetic neurons can be generated in graphical form that are both visually and statistically indistinguishable from equivalent populations of biological neurons selected by cytoarchitectural area, layer, age, and neurological diagnosis from the developing human cerebral cortex. Representative populations of several neuron types (pyramidal, spiny stellate, basket cells, chandelier cells, etc.) have been modeled. The long-range goal of these studies, anatomical realism, transcends the traditional statistical view that models the neuron database exclusively as a set of parameters and their variances. Anatomical realism requires that sufficient parameters be captured that a credible forest of neurons can be grown synthetically within the computer and displayed graphically.