Brain-machine interfaces as the new frontier in extreme miniaturization

The exact functioning and operation of the brain has been and still is to a major degree a great mystery. The recent introduction of advanced imaging tools such as fMRI, EEG and eCoG and, most recently, direct neural sensing are throwing the doors of neuroscience wide open, and enable direct in-vivo observations of the brain at work in dynamic conditions. This may help to address a broad range of neural impairments and diseases, such as stroke, paralysis, epilepsy, depression, etc. However, for all of these to happen it is essential that neural interface circuitry is developed that surpasses the state of the art in ultra-low power miniaturized design by at least an order of magnitude. Furthermore, the resulting sensory/stimulation nodes have to be energy-self contained and support wireless links > 1 Mbps. This paper explores the opportunities of accomplishing just that, and demonstrates the feasibility with a number of examples. The potential outcomes of these developments are just "mind-blowing", and can dramatically impact the evolution of human-cyber interfaces in the decades to come.