Multi-Level Neuron and Network Modeling in Computational Neuroanatomy

Abstract

While quantitative neuroanatomy produces increasingly detailed descriptions of the nervous system at all levels of organization, it remains a major challenge to integrate the information from cell, population, and systems levels in a mutually informative way. Computer simulation is a powerful tool for exploring links between structure and function of cells, tissues, and organs. It is based on mathematical models that capture the essence of anatomical, physiological, and behavioral observations, and rely on accurate quantitative descriptions. Previous computer models, including much of our own work, have focused on relating structural and functional data at nearby descriptive levels. Here, we discuss concepts in linking cell, population, and systems level models of the cerebral cortex. We propose a strategy for building multilevel computer models that integrate elements from compartmental neuron models, microcircuit representations of neuronal populations, and activity propagation in large-scale neuronal networks. As a working example, we simulate activity propagation in the primate visual cortex, with the aim of relating neuronal activity to cortical activation patterns and onset response latencies to the structure of the underlying anatomical network. This computational approach provides new insights into the functional anatomy of the visual cortex.