Mesoscale holographic interrogation of visual circuits mediating predictive coding

ended

Predictive coding is an attractive framework to understand sensory processing. However a direct experimental demonstration of this unique capacity of the brain to sensory representations has been lacking, essentially because of the difficulty to separate between external and self-generated representations using correlative or causal paradigms. Similarly, the neural circuits underlying predictive coding in the context of sensory perception have been largely unexplored, primarily because of technological bottlenecks to image and manipulate selected neurons within neural circuits extending multiple areas. In this proposal, I aim to develop a mesoscale two-photon holographic technology that will enable the manipulation of functionally selected neurons in targeted higher visual areas, while simultaneously measuring neural activity throughout the cortical visual hierarchy. This will allow to test primary hypotheses related to the predictive coding framework in the context of visual processing. Importantly, it will provide the possibility for the direct demonstration of predictive coding in the visual cortex by artificially generating an internal sensory representation from scratch in a behaving awake animal. Finally, I use will use this powerful technology to investigate the neural circuits underlying top-down predictive coding with a precision and scope unattainable so far.