Recording brain activity at the mesoscopic scale has a strong potential to unveil many new fundamental neuronal operations. Optical imaging offers a unique opportunity to measure brain activity over a large area with high spatio-temporal resolutions (20 μm x 1 ms). However, two major limitations of this imaging technique partially explain the lack of development in this field. The cortex being non-planar, the field's depth limits the region in focus to a small region close to the center of the field of view. This is particularly significant for the highly curved lissencephalic small cortex of non-human primates that are becoming popular in neuroscience experiments. The ideal technique would be a method that compensates for such curvature; it would enable imaging the whole visual system at once, from the primary to the fifth visual cortices, in small non-human primates. Additionally, the signal-to-noise ratio is strongly degraded by the dynamic evolution of the brain curvature due to physiological rhythms (heartbeat, breathing, etc.). This strongly limits the ability to work at a single-trial level and to unravel the real dynamics of neuronal processing, such as spatio-temporal waves. Here in this project, we present an interdisciplinary approach for imaging of the non-human primate cortex, using technologies from astronomical instrumentation to overcome current technological limits. This will be of interest to a wide neuroscientific audience but also will impact the clinical community interested in mapping the nervous activity at the mesoscopic scale. Our current preliminary development involves redesigning the illumination source and the optical design.