Home

About

Research

Events

Contact

print

Banbury Center Workshop Series

Neurocomputational Strategies: From Synapses to Behavior (1998)
Functional Organization of the Thalamus and Cortex and their Interactions (1999)
Toward Animal Models of Attention and Consciousness (2000)
Persistent Neural Activity (2000)
Can a Machine Be Conscious? (2001)
Communication in Brain Systems (2004)
Neurobiology of Decision-Making (2005)
Computational Approaches to Cortical Functions (2006)
New Frontiers in Studies of Nonconscious Processing (2007)
Theoretical and Experimental Approaches to Auditory and Visual Attention (2008)
Searching for Principles Underlying Memory in Biological Systems (2009)
Neuronal Response Variability and Cortical Computation (2011)
Connections and Communication in the Brain (2014)

Michael P. Stryker

Simple and Complex cells in a large-scale neuronal network model of primary visual cortex

University of California, San Francisco

Two major approaches have informed our thinking about the way the world is represented in the brain. One approach used in both humans and experimental animals has been to make maps that show how stimuli are represented across the cortical surface. In their simplest form, brain maps are unassailable. Visual stimuli, however, have many qualities, including retinal position, shape, motion, binocular depth, color, texture, and others. Any actual stimulus has a particular combination of these qualities. Both conceptual and technical difficulties for brain maps arise when they are made to represent multiple qualities of a stimulus, as illustrated in recent experiments on the representation of orientation and spatial frequency in primary visual cortex (Issa et al, J Neurosci 20:8504, 2000). Conventional methods of brain mapping incorporate two sets of false assumptions: first, an assumption of something like linearity, that the maps of each stimulus quality can be obtained by averaging responses over different values of the other qualities; and second, an assumption that the average of the range of stimuli used in any particular experiment is an appropriate baseline for normalization of responses.

The other major approach used to study how the world is represented in the brain is to characterize the response properties of individual brain cells. This characterization is referred to as the receptive field of the neuron. It consists of determining the values of each stimulus quality that produce the strongest response from the neuron, and of plotting the curves that show how the neuron's response declines as the values of each stimulus quality are changed from the optimal.

We suggest that brain maps have a mathematical dual in the concept of the receptive field. In principle, one should be able to go back and forth between these dual notions of receptive fields and maps by changing focus from the overall pattern of brain activity in an area to the properties of the individual element that is responding. Until now, this duality has not been exploited in mapping cortical organization. We argue that the appropriate framework for the mapping of multiple stimulus qualities in a brain region is a determination of the characteristic receptive field at each point on the brain.