Edinburgh Research Explorer

Neural Maps: Their Function and Development

Research output: Chapter in Book/Report/Conference proceedingChapter

Original languageEnglish
Title of host publicationFrom Neuron to Cognition via Computational > Neuroscience
EditorsMichael Arbib, James Bonaiuto
PublisherMIT Press
Pages409-432
Number of pages34
ISBN (Electronic)9780262335263
ISBN (Print)9780262034968
Publication statusPublished - Nov 2016

Abstract

In this chapter, we first define a neural map as a sheet of neurons systematically related to another population of neurons, including the special but common case of a topographic map as a spatially organized mapping of this type. We further define a feature map as a topographic mapping of some underlying features or patterns in the input data, beyond the strict anatomical arrangement of the input region. Next, we review the main biological findings about the properties of these maps in adult animals, which will be expanded in more detail when discussing specific models of development and function later. We use well-established examples from the visual system, but highlight similarities and differences with maps in other sensory modalities and in motor regions.The remaining sections present models and analyses addressing a series of key questions about neural maps:Where do neural map patterns come from? How do feature maps arise from neural mechanisms? What is the information-processing goal of neurons in maps? Together, the models and data suggest that systematic maps first emerge in sensory areas through genetic specification of initial topography, based on molecular guidance cues. The maps are then shaped by activity-dependent and experience-dependent processes in neurons and their connections that lead a neural population to have good coverage of possible stimuli,with neurons responding to the full range of stimuli encountered during development. These processes also typically result in locally smooth maps, though the functional significance of this organization remains unclear.

ID: 23784591