New brain implant tech from Blackrock is making ‘mind over matter’ a reality

In 2012 quadriplegic Jan Scheuermann had two electrode arrays implanted in her brain’s left motor cortex. Within a week, she could use the implant to move a prosthetic arm in all three dimensions, and move its wrist in an additional four. That was good enough for her to use it to take a sip of coffee with the arm, and good enough to make her a national celebrity. Researchers from the University of Pittsburg have now been able to finesse an additional three degrees of freedom out of an improved implant-to-arm pipeline by swapping in additional finger motions where previously Jan had only a crude pincer grip.
It is hard to underestimate the difficulty, and therefore value, of making advances like this to the field of brain-computer interfaces or BCIs. Mapping specific functions to specific places has generally proven difficult. We know, for example, that with crude TMS stimulation, the most reliable response obtained for activating an cortical motor area discovered to date is a simple adduction of the thumb (bringing it back into the plane of the palm of the hand). In fact, that is precisely what is used as the reference point to calibrate the coil’s position and power. It is no surprise then that researchers have now been able to add more subtle thumb control to the existing suite of prosthetic powers.

To be more specific, the new functions they describe are finger abduction (spreading), thumb opposition, finger and thumb pinch (which also happens to be an important gesture for communication), and something a little more esoteric we can just call “the scoop.” The video above gives some indication of the speed and accuracy that is now possible. The science of mapping signals recorded from electrodes at various locations, to functions, as it is now practiced, doesn’t actually use any special knowledge of anatomy or physiology other then the initial placement of the array. The signal processing techniques that they use, which we have described in more detail in our post about bionic eyes, have a long history and are only now reaching the point of doing something useful.
To move forward researchers will need to supplement these methods with better anatomical information about how to best distribute multiple arrays in different regions of the brain. A major concern at the moment is making these arrays safe and legal. To that point Blackrock Microsystems has just established an overseas location in Hanover, Germany where some of the legislation on thought-controlled prosthetics is not as overbearing. In the time since the video was made, Jan’s had her own electrodes removed — a procedure that can’t be fun for the brain. Other researchers are looking to make implants from biodegradable materials that disappear on their own. While these kinds of devices are nowhere near the ability to do what 100-electrode microarrays can now do, they can potentially address of the more practical issues of longevity and device failure procedures that has the FDA concerned.

The complete implant, with the electrode array at the bottom, and various other parts above it. Only the electrode array is actually inside the skull.

Blackrock not only manufactures the electrode arrays but also much of the peripheral equipment that is needed — electrical feedthroughs, connectors, software, and hopefully soon, wireless capability. Blackrock recently brought engineer Ming Yin from Brown University on board, and also licensed much of the technology described in his groundbreaking paper in the latest issue of Neuron. As shown above, Yin’s new design mates a Blackrock electrode array to a wireless link which is capable of transmitting data obtained at high sampling rates (20,000 samples per second per channel) up to a distance of 5 meters. The ultra-low-power, 46-gram, custom ASIC-enabled device can operate continuously for 48 hours. That’s enough for some serious coffee drinking.
With companies like Blackrock providing off-the-shelf solutions, and universities around the globe working to perfect signal processing software, we should soon get a better handle on the possibilities, and limits, of this approach to BCIs. With electrodes that can reach ever more remote recesses of the brain, the hope is there will be less guesswork and more precision in extracting intent from the mind.