Functional connectivity within the motion system of the macaque monkey

Motion processing in primate cortex is organized hierarchically. Both physiology and motion processing theory start small and local, and progress consecutively to larger and more global representations in higher cortical areas. Receptive fields (RFs) at each level of the hierarchy are constructed from their precursors.;Within the macaque motion processing hierarchy, the middle temporal area (MT) is the first area containing a high concentration of direction-selective cells (> 80%), and is one of the best-understood extrastriate visual areas. Of the areas showing sensitivity to global motion patterns, the medial superior temporal area (MST) has received the most attention. Both areas have dense feedforward and feedback connections between them; for these reasons we also used MT and MST to study functional connectivity, MT and MST RF formation, and the function of feedforward and feedback connections within the motion system. Our studies measured the integration and segregation of motion signals for the formation of the MT and MST RFs. It was the larger goal of these studies to determine how and where these two important aspects of brain organization come into play and balance one another in the motion processing system.;We used cross-correlation analysis to measure the functional connections within area MT and between MT and MST. We were also particularly interested in measurements of functional excitation and inhibition, as the processes of integration and segregation are mediated by such activities. The MT and MST cells we recorded from could have very similar or very different neuronal properties, allowing us to measure both integration and segregation processes for the formation of their RFs. We were interested in understanding how integration of directionally selective inputs balances with the segregation of non-preferred motion directions. In addition, we measured dynamic changes in these connections while varying stimulus location and motion direction.;We found strong functional excitation between MT neurons sharing similar tuning properties and RF locations. This integrative connectivity was stronger when the stimulus overlapped the RF locations and moved in the cells' preferred direction. In contrast, MT-MT and MT-MST neuron pairs with opposite tuning preferences showed functional inhibition. These are exciting, new findings and may reflect the segregation and suppression of unwanted motion representations in the non-preferred motion direction. Within MT, correlation depended strongly on stimulus direction and was spatially localized to the MT RF location. This was not the case for correlation between cortical areas MT and MST, indicating a more spatially and directionally specific circuit within MT.;Finally, we used adaptation to perturb MST feedback to MT. We used a large expanding stimulus outside the MT RF specifically designed to adapt area MST neurons without directly affecting the MT cell under study. Thus, we tested the possibilities of both high-level adaptation and feedback adaptation effects, and found that adaptation outside the RF affected the MT cell's response. This adaptation could have been caused by adaptation of MST neurons and the effect passed to MT in a feedback manner.