THALAMIC CONTROL OF CORTICAL VISUAL PROCESSING

点击翻译按钮获取中文摘要

The long-term objective of this project is to understand how afferent parallel channels contribute to the processing of visual information in cerebral cortex. The output of the retina is organized into at least three channels (W, X and Y cells), which are relayed through the lateral geniculate nucleus (LGN) to the visual cortex. The LGN is a laminated structure, and the pattern of lamination varies greatly among species. In the cat, three retinal channels are distributed among nine layers of the LGN complex, creating at the morphological level of organization several additional afferent channels. Many individual cells in cortex receive convergent inputs from multiple channels through both feedforward and feedback pathways. Cortex and other central structures project heavily to the LGN, and therefore are capable of strongly regulating LGN transmission to cortex. The complex interactions of afferent channels, intracortical circuits, and central feedback to the LGN are probably dynamically regulated in accordance with the perceptual demands and behavioral state of the animal. Examining these circuits in the awake, behaving animal is critical for insights into their functional roles. Three aspects of these interactions will be examined in awake cats trained in visuomotor tasks. Cells in the LGN will be recorded to investigate the effects of saccades, gaze angle, and spatially-selective attention on their activity, and to determine how these effects vary by layer and by cell type. The activity of corticogeniculate cells in primary visual cortex will be examined under the same behavioral circumstances to identify dynamic changes that are common to or different from those observed in LGN cells. In order to understand the contributions of individual LGN layers to dynamic changes in cortex, layers will be selectively inactivated with microinjections of blocking agents while visuomotor behavior and cortical activity are observed. To gain insights into the origins of interspecies variations in LGN laminar structure, the morphogenesis of the monkey LGN will be model led with simulated annealing techniques. The aim is to test the hypotheses that in the rhesus monkey, eccentricity-related variations in the number of layers are promoted by regional variations in retinal ganglion cell density, and that the blind spot determines the point at which the pattern changes by serving as a "seed crystal", causing an abrupt change in the anterior-posterior free-energy gradient that determines the most stable state.