23 August 2015

Stanford researchers build fully internal optical brain interfaces

Two prominent Stanford researchers, Ada Poon and Karl Deisseroth, recently teamed up to create a completely wireless optogenetic implant. Instead of relying on fiber optic tethers and bulky headset receivers, their tiny mouse stimulator generates light from LEDs that are powered with an ingenious technique: a 1.5 GHz RF cavity that couples energy to the implant by using the whole mouse as an efficiently matched receiver.

In contrast to more conventional inductive energy transfer systems that need to have direct coupling between two opposed coils, the animal is free to move about anywhere above the energizing lattice in the floor of its chamber. But this is not just some scaled-down version of a subsurface highway charger for electric vehicles. Instead, resonant excitation of a confined electromagnetic field pattern (i.e. its intrinsic mode) can be localized to the mouse independent of its position.

It’s all in a closed-access paper Poon previously published not so long ago. We are not yet sure how to scale this up to humans. But as long as our dielectric properties are similar, the main variable should just be physical dimension. Provided you get that right, and have a way to get a few opto-enabled ion channels preferably the channelrhodopsin 2 (ChR2) variety into select parts of your nervous system, the actual hardware to rectify sufficient power for the LEDs is fairly simple.

In fact, all you need to boost up the raw DC voltage is four Schottky diodes and four capacitors configured into a two-stage doubling circuit. Together with three small turns for the antenna receiver, everything should fit into 10 mm3 package that weighs under 20mg. As the researchers show in their actual experimental paper, that’s small enough to fit nearly anywhere in the central or peripheral nervous system, even just under the skin at sensory nerve endings.

Just to make sure everything is on the up and up, you may initially want to use a bit of exploratory fiber optics anyway. They make for a highly accurate and localized temperature probe. In the course of due diligence, the researchers demonstrated that any incidental temperature rises associated with the stimulation were limited to <1 °C. That presumably includes any heat from LED light itself, the associated electronics, and the more generalized absorption of RF energy. More sophisticated power conversion might even wring a little more efficiency from the micro LEDs. The researchers estimated they were getting about 20% efficiency (emitted light power/input power) while the manufacturer spec sheet indicated that efficiencies up to 60% should be possible. More importantly perhaps, the researchers could generate light pulses as tight as 100 μs. This kind of temporal precision would allow one to exploit the full dynamic range of the channel opsin now available. We have been chronicling the dramatic advances made by both Poon and Deisseroth for several years now. The technological fruits now falling out of their mutual labor, while clearly awesome just to behold, are even more seductive when we can fully view them in their transparent simplicity with an eye to one day possess.



Post a Comment

Get every new post delivered to your Inbox.


Copyright © 2015 Tracktec. All rights reserved.

Back to Top