Student Theses and Dissertations

Date of Award


Document Type


RU Laboratory

Gilbert Laboratory


primary visual cortex, association field, cortical reorganization, visual perception, v1 neurons


The ability to derive meaning from complex, ambiguous sensory input requires the integration of information over both space and time, as well as cognitive mechanisms to dynamically shape that integration. We have studied these processes in the primary visual cortex (V1), where neurons have been proposed to integrate visual inputs along a geometric pattern known as the association field (AF). We first used cortical reorganization as a model to investigate the role that a specific network of V1 connections, the long-range horizontal connections, might play in temporal and spatial integration across the AF. When retinal lesions ablate sensory information from portions of the visual field, V1 undergoes a process of reorganization mediated by compensatory changes in the network of horizontal collaterals. The reorganization accompanies the brain’s amazing ability to perceptually “fill-in”, or “see”, the lost visual input. We developed a computational model to simulate cortical reorganization and perceptual fill-in mediated by a plexus of horizontal connections that encode the AF. The model reproduces the major features of the perceptual fill-in reported by human subjects with retinal lesions, and it suggests that V1 neurons, empowered by their horizontal connections, underlie both perceptual fill-in and normal integrative mechanisms that are crucial to our visual perception. These results motivated the second prong of our work, which was to experimentally study the normal integration of information in V1. Since psychophysical and physiological studies suggest that spatial interactions in V1 may be under cognitive control, we investigated the integrative properties of V1 neurons under different cognitive states. We performed extracellular recordings from single V1 neurons in macaques that were trained to perform a delayed-match-to-sample contour detection task. We found that the ability of V1 neurons to summate visual inputs from beyond the classical receptive field (cRF) imbues them with selectivity for complex contour shapes, and that neuronal shape selectivity in V1 changed dynamically according to the shapes monkeys were cued to detect. Over the population, V1 encoded subsets of the AF, predicted by the computational model, that shifted as a function of the monkeys’ expectations. These results support the major conclusions of the theoretical work; even more, they reveal a sophisticated mode of form processing, whereby the selectivity of the whole network in V1 is reshaped by cognitive state.


A thesis presented to the faculty of The Rockefeller University in partial fulfillment of the requirements for the degree of Doctor of Philosophy.

Permanent URL

Included in

Life Sciences Commons