Scientists at Brown University have developed a fully implanted, wireless, biologically safe neural interface. A major scientific milestone in research toward effective brain-computer interfaces, the device was primarily designed for recording motor activity from the brain. In practice the device could be used by a patient to do things like control a robotic prosthesis or electronic mouse cursor without needing to be wired directly to laboratory equipment. Copies of the device have been implanted and operating in pigs and non-human primates for more than a year as part of a test of it’s operation.
On the outside the implant is fairly nondescript, a compact titanium rectangle with a small glassy blue window set into the top side, with a fine electronic lead exiting from the bottom. Internally though, the implant is a small engineering marvel. Within its 56 x 42 x 9mm housing (completely hermetically sealed to prevent biological contamination or leakage) Brown researchers packed a complete signal processing system: a tiny microelectrode array that reads neuron output, lithium ion battery, inductive coil for wireless recharging, radio/infrared transmitters, and purpose-designed low power circuits to convert the neural signal into a digital format.
“What makes the achievement discussed in this paper unique is how it integrated many individual innovations into a complete system with potential for neuroscientific gain greater than the sum of its parts,” said David Borton, lead author of the paper. To him what makes this such an important advance is not just furthering our knowledge of the brain, but also the potential for a clinical translation of this or subsequent similar BCI’s. “Most importantly, we show the first fully implanted neural interface microsystem operated wirelessly for more than 12 months in large animal models — a milestone for potential [human] clinical translation.”
Though is not approved for clinical trials in humans at this time, the Brown team attests that it was designed with potential translation in mind. Before that happens there are a few engineering challenges yet to be overcome, although those are now more a matter of refinement than innovation.
The device requires two hours to charge and thereafter can function continuously for seven hours, transmitting neural information at 24 Mb/s to an external receiver.
The result of the design/study of the device was announced last week at the 2013 International Workshop on Clinical Brain-Machine Interface Systems in Houston; the paper describing it is published in the Journal of Neural Engineering. Meanwhile, scientists at Brown are continuing to work on advancing the device further by making it smaller, lower power, and more robust.