Towards an understanding of the synergistic properties of cortical processing: A neuronal computational modeling approach

The overarching goal of cognitive science research is to understand the organizing principles of brain physiology that underlie human behavior. To-date, neurobiological studies have focused on single neuronal pathways (especially vision, touch, and hearing) yielding tremendous knowledge of primary subcortical and early cortical processing, but limited insight into the sensory integration needed for higher-level processing. It is my hypothesis that if we simultaneously study multiple sensory pathways, we should be able to cull out the synergistic properties common across modalities, thereby bringing the science a step closer to understanding the neural correlates of high-level cognition.; In fact, daily human interaction with the environment regularly entails assimilation of information using multiple sensory systems. Integration of multiple modalities is especially beneficial when a single sensory stream provides low signal-to-noise or an ambiguous interpretation. Touching while observing, tasting and smelling, listening while lip reading are common examples. A classic example is the “cocktail party problem”, whereby listening to, and reading the lips of a nearby person allows one to override the loud superposition of nearby voices. These two modalities-vision and hearing-are particularly appealing as candidates for simultaneous study because they independently mediate the highest of cognitive functions: human language.; But how can one approach this complex challenge? Cognitive science attempts to explicate brain processing using symbolic representations of our environment with a top-down approach. It is unclear if the brain utilizes such symbolic representations. Information processing models, popular among cognitive scientists, assume cognition can be understood in terms of a series of sequential stages with each stage providing unique processing leading to an output. In contrast, neuroscience takes a bottom-up approach to understanding cognition; for example, anatomical and electrophysiological studies have shown that reciprocal and parallel connectivity between functionally independent neurons is the rule rather than the exception. The general approach of my dissertation will be to develop a spiking neuronal network of bimodal processing using the latest information provided from biology and neuroscience to determine the synergistic properties and neural correlates of multimodal processing.