In the pursuit of restoring vision for patients having lost the eye-brain connection, ultrasound stimulation is a new interesting therapeutic approach at the cortical level. In particular, Sonogenetic therapy has been explored as a mean to increase the sensitivity of cells to ultrasound by expressing a mechanosensitive ion channel in the membranes of cells of interest in order to modulate their activity.In order to optimize sonogenetics for vision restoration, we first focused on exploring the properties of different mechano-sensitive ion channels, the key parameters being the spatial and temporal resolutions. We tested and compared mechano-sensitive ion channels such as TRPA1 in order to pinpoint the most suitable one. For example, our results on eMscL-G22S suggest that our approach to sonogenetic therapy could yield a spatial resolution under 400 μm for stimulations at 15 MHz and very short latencies, namely, 12.2 ± 2.5 ms.As our goal is to target the neurons of the primary visual cortex (V1) to activate them with focused ultrasound, we injected Non-Human Primates in the V1 area with a viral vector to express eMscL-G22S. First, we studied V1 to understand its functional architecture. Then we investigated behavioral response to sonogenetic stimulations through saccade tasks. Our next step is to create complex shapes (letters, geometric shapes) and to define an optimized sonogenetic stimulation strategy.This will allow us to develop a therapeutic approach good enough to restore usable vision for blind patients.