Communicating with neurons at the speed of light

Advances in brain–machine interfaces (BMIs) have been driven by a proliferation of neural sensing and stimulating modalities, as well as an increase in the number of neurons that can be simultaneously recorded or stimulated. These advances have led to new insights into neural systems and the development of novel clinical tools. Ideally, a BMI should be minimally invasive, extremely safe and offer sufficient longevity. It should have a feedback system to close the communication loop and have high spatiotemporal resolution and throughput. Imaging and stimulating neurons optically, rather than electrically, holds tremendous promise due to the cellular selectivity and high spatial and temporal resolutions of the technique. Computer-Generated Holography (CGH) using Spatial Light Modulators (SLMs) can sculpt light to address ensembles of neurons simultaneously, however available SLM speeds constrain the bandwidth and throughput of such interfaces. Additionally, high CGH compute times limit the ability to provide the real-time computations required to close the loop. We propose a high-speed Microelectromechanical Systems (MEMS) based SLMs capable of switching at kilohertz speeds coupled with a high-speed CGH algorithm that can open the door to closed-loop all-optical neural interfaces with high spatial and temporal precision.