Brain-Machine Interfaces as a new non-pharmacologic approach for treating brain disorders


                More than twenty years ago, a decade long collaboration between John Chapin’s group at Hahnemann University, and my lab at Duke University culminated with the first experimental demonstration that a direct link between animal brains and machines could be established and maintained for long periods of time.1 This pioneer demonstration of a brain-machine interface (BMI), as the new approach became known, was only possible because of a novel neurophysiological technique, developed and perfected by our two labs.2,3 By the late 1990s, this technique, known as chronic, multi-site, multi-electrode recording, took advantage of permanent implants of multiple arrays of flexible microelectrodes to allow the simultaneous recording of the extracellular electrical activity of 50-100 individual neurons in freely behaving animals.4,5 In addition to remaining viable for several weeks or even months, these chronic brain implants allowed us to record single neurons located in multiple cortical and subcortical structures in the same subjects, hence creating the opportunity to monitor entire neuronal circuits in freely behaving animals for the first time. Using this method, my laboratory soon demonstrated that both New and Old World monkeys could effectively operate a BMI.4,5 By 2004, the viability of directly linking the human brain to artificial devices was also demonstrated in human patients and later on replicated by several other groups.6,7,8 In parallel to these invasive BMIs, non-invasive methods like scalp electroencephalography (EEG), were employed to demonstrate for the first time the operation of brain-computer interfaces, initially in patients suffering from locked-in syndrome.9