Functional circuits in visual cortex: multi-photon imaging and computer modelling. (360G-Wellcome-076694_Z_05_A)

It has long been known that the retina adapts to the prevailing level of lightentering the eye. It is also now clear that the visual cortex can dramaticallychange its profiles of temporal and spatial integration depending on the pattern, the movement, and the contrast of the scene. A striking example is adaptive temporal integration (ATI), which Movshon and I discovered in complexdirection selective cells of the primary visual cortex (V1): the responses of these neurons speed up as they are challenged with faster moving stimuli. Findings of this type have far-reaching implications for mapping and modellingneuronal function, particularly for understanding how the visual cortex responds during natural viewing, when the statistics of the scene change frequently from moment to moment. The key goal is to integrate computational modelling and experimental electrophysiology into a unified research group to formulate a biologically plausible model that explains ATI and other fast-acting adaptive phenomena. Major cell classes in V1 will be tested with dynamic stimuli to find the origin of ATI. The spatial and temporal selectivity of ATI and its influence on synchronous firing will be probed in V1 and in the motion area V5/MT. The resulting model will adhere to four principles: (1) modular components will represent major cell classes at several levels from the retina to the perceptually relevant responses in V5/MT, (2) arbitrary time-varying stimuli will be accepted as input, (3) responses will be given as spike trains and membrane voltages, (4) the model will be publicly available online so that it can be utilized and refined by scientists anywhere in the world.