Computer modeling will be used to elucidate the mechanisms that underlie direction selectivity (DS) in simple cells of primary visual cortex. DS serves as a good model for understanding the more general problem of how the brain performs behaviors that require precise temporal information. This problem is highly relevant clinically; for example the disruption of temporal processing is an important factor dyslexia. In visual cortex, a neuron's DS derives from inputs that have their own spatial and temporal structure. These inputs may arise from three sources: the lateral geniculate nucleus (LGN) and other excitatory or inhibitory cortical cells. A key problem this study will address is to identify those spatiotemporal relationships among LGN afferents and interacting cortical neurons which initially create or enhance DS. Major project goals will be to: (1) Develop a computer model that investigates the ability of lagged and nonlagged LGN cells alone to establish a directional preference in area 17 simple cells. (2) Incorporate into the model intracortical excitatory and inhibitory connections and examine their ability to establish or enhance a directional preference. (3) Simulate the development of direction selectivity through Hebbian learning. (4) Evaluate the validity of the computer model by comparing the response of model cortical units with those of actual cat area 17 simple cells for a variety of test stimuli.
