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)

Tony Movshon

Local And Global Operations Performed By Neurons In Macaque Area MT

New York University

Neurons in macaque area MT have receptive fields that are roughly 30 times larger than those of their inputs from V1. It is generally assumed that the large RF size in MT results from convergence of inputs from a wide area of V1, but the way that visual signals are transformed by that convergence is unclear. We have uncovered two contrasting rules for this convergence, one local and one global.

One observation is derived from the demonstration that MT neurons combine information from multiple orientation sensitive neurons in V1 to compute "pattern motion". When the different orientations that define pattern motion are presented to spatially separated patches within the RF, the pattern motion calculation is disrupted and MT neuron responses mostly reflect the independent local motions.

A second observation is based on earlier results showing that cortical neuron responses are reduced following a period of adaptation to a strong stimulus. When adapting and test stimuli are presented to different parts of an MT RF, the adaptation effect does not transfer from one patch to the other, suggesting that the adaptation effect, like the motion calculation, is based on local computations.

A third observation is based on the observation that visual stimuli affect the contrast gain of cortical neurons through the action of a recurrent gain control network. Presenting a stimulus to one part of an MT RF strongly affects contrast gain for stimuli presented in other, remote parts of the RF, suggesting that contrast gain control is based on global, not local, computations.

These observations are consistent with a model in which most RF properties are determined by local circuits before they converge to create large RFs. Overall contrast gain, however, seems to be different, and is controlled by convergent signals from the whole large RF.