Developing methods to interface directly with the cortex through targeted stimulation is vital not only for probing fundamental neural mechanisms, but also for developing clinical solutions such as visual prostheses. Currently, the standard method for cortical stimulation in primates relies on intracortical electrical stimulation using either rigid or flexible electrode arrays. However, these traditional approaches face several fundamental limitations. First, their channel count and stimulation density — constrained by their inherent invasiveness — fall short of what is required to accurately manipulate the primate cortex. For example, the primate primary visual cortex (V1) is organized into a dense mosaic of functional domains with a diameter of just 0.1 to 0.2 mm, each tuned to process specific visual features (such as colour or orientation) at a particular location in the visual field. Precise targeting of these domains requires a density of at least 100 stimulation sites per square millimetre, which would be likely to severely damage the brain if intracortical electrodes were used. Furthermore, electrical stimulation indiscriminately activates all neurons and passing axons that are close to the electrode, lacking the ability to distinguish between even the most basic types of excitatory and inhibitory neuron. Thus, it is apparent that there is a crucial technological gap: a minimally invasive, high-density and cell-specific tool to precisely control the primate brain is needed.

To address this challenge, we developed an approach for high-resolution stimulation of the primate cortex using mesoscale optogenetics. When selecting this approach, we drew inspiration from everyday technology. The artificial devices that have the largest number and highest density of independent output channels are, arguably, optical displays. For example, the pixel count of a modern mobile phone display easily exceeds the roughly one million fibres in the human optic nerve. Furthermore, unlike physical probes, light is non-invasive and can be projected into the brain without causing mechanical damage. We asked: why not use a high-resolution optical display for cortical stimulation?