Skip to main content

Computations by Neurons and Populations in Visual Cortex

Final Report Summary - CORTEX (Computations by Neurons and Populations in Visual Cortex)

This project explored the computations performed by neurons and populations in the cerebral cortex. We focused on the part of the cortex that is concerned with vision, the “visual cortex”, and we studied it in the mouse, as this is a species that allows one to perform highly controlled experiments in situations that mimic natural behaviour. These include setups for virtual reality, in which the mouse explores a virtual environment while we record from individual neurons and populations. The project involved a number of techniques, some established and some new. Established techniques included intracellular recordings, where one inserts a glass pipette in the brain and records the internal voltage of individual neurons, and multielectrode recordings, where one inserts an array of electrodes and records the voltage in the immediate vicinity of a number of neurons to track their activity. New techniques involved imaging from genetically-engineered mice whose brain cells expresses fluorescent proteins. The fluorescence of these proteins tells us about the voltage of the cells. Using these techniques we made a number of discoveries. For instance, we discovered that there is a fundamental difference in the role of inhibition in the cortex depending on whether the cortex is awake or anesthetized. All previous measurements had been made under anesthesia, so this result was unexpected and unknown. We also discovered that some fundamental attributes of neurons in visual cortex, namely how they integrate images over space, and how they process visual motion, are in turn shaped strongly by how an animal explores an environment. This was not known, as usual measurements done on these cells were made while animals were stationary or not navigating an environment. We also discovered the rules that govern communication between neurons located in disparate locations in cortex: these rules can be summarized by a simple arithmetical expression: addition when overall activity is low, division when overall activity is high. Taken together, these advances constitute a success of the project, one that we hope that our laboratory and other laboratories will be able to build on in future years.