Adaptation and optimization of information processing in the early visual cortex

The brain is arguably the most powerful and complex computer in the world. The broad aim of the proposed research program is to understand the types of calculations that are performed by the brain. This research focuses on parts of the brain responsible for vision because the visual stimuli that serve as inputs to the system can be precisely controlled. Although the eyes detect light, it is the brain that does most of the work that allows us to see the world around us. Millions of brain cells, also called neurons, work together to analyze the visual scene that is projected onto the back of the eye. It has been known for a long time that neurons located in a part of the brain called the primary visual cortex detect specific features such as lines or edges in the visual scene. Interestingly, the preferred stimulus of these neurons can change in a context-dependent manner. What surrounds a feature in space and what has been observed in the recent past both affect how neurons in the primary visual cortex respond to a visual stimulus. The proposed research will use electrophysiological methods to record from individual neurons in the primary visual cortex of cats to clarify the mechanisms and function of these context-dependant changes. Furthermore, the responses of these neurons will be examined using formal information theory to asses whether context-dependent changes could improve the accuracy or efficiency of visual processing. Finally, human psychophysics will be used to investigate how the context-dependent changes presumed to also occur in human primary visual cortex affect visual perception. The results of these studies will be fascinating from an information processing perspective because a stable visual reality is maintained despite the fact that individual neurons continuously adjust their response properties. The proposed research will also contribute to the development of more accurate computer simulations of the primary visual cortex thereby aiding our understanding of early cortical coding, and may serve as inspiration to biomimetic models of computer vision.