February 28, 2013 | 3
Miguel Nicolelis is a brilliant neuroscientist (and showman) who is constantly trying to explore how far technology that uses brain signals to control machines can be pushed. In his 2012 book Beyond Boundaries, he speculated about an experiment in which two rat brains would exchange information—telepathic tweets, if you will.
He wrote in one chapter: “In this arrangement, could the two brains eventually reach a consensus, let’s say, about the identity of a complex object explored only partially by each rat? Would the rats literally share their minds to build vicarious sensations, through a sort of touchless, Vulcan mind-melding ritual, in order to overcome the limits of their individual brains?
Nicolelis just took one small step toward a brain net. He reported in the Feb. 28th Scientific Reports (part of the Nature Publishing, as is Scientific American) that his research team at Duke University Medical Center had achieved a back-and-forth exchange between two rodent brains. Not exactly a Vulcan mind meld, but the experiment demonstrates that the technology of brain-machine interfaces—which relays a signal from the cerebral cortex to a prosthetic—might be extended to a transfer of signals from cortex to cortex.
Would this enable a true meeting of minds—and would that be better than a chat over a latte at Starbucks? Nicolelis has speculated that these interfaces could eventually lead to networks of interconnected cocos that would lend new meaning to the idea of a brainstorming.
Still a ways to go until brain nets. But here’s what happened: both animals were trained to press one or the other levers when an LED turned on in exchange for a drink of water. Microelectrodes were then placed in each of the animals’ cortices and when one rat pressed the correct lever, a sample of cortical activity was wired to the second in a second chamber where the “it’s-time-to-drink” LED was absent. The rat on the receiving end, nonetheless, proceeded to press the correct lever that had been messaged over the brain link. It did so an average of 64 percent of the time, compared to chance (50 percent) and an accurate hit rate of 96 percent for the rat sending the signal after witnessing the LED illuminate. A similar experiment was attempted in which one rat communicated the presence of a narrow or wide opening to the other. One of the experiments was performed with an over-the-Internet link from Natal, Brazil to Duke in Raleigh.
The intensely competitive brain-machine interface community is chary with compliments. And Andrew Schwartz, another luminary in the field, was less than wowed. He told Ed Yong for Nature News: “Although this may sound like ‘mental telemetry’, it was a very simple demonstration of binary detection and binary decision-making. To be of real interest, some sort of continuous spectrum of values should be decoded, transmitted and received.”
One source suggested by Nicolelis to reporters was more generous. Marshall Shuler, a neuroscientist at Johns Hopkins, registered high praise: “This work further advances a large body of inquiry conducted by the Nicolelis lab over many years, –specifically by closing the “control loop” for brain actuated technology– in pursuit of Miguel’s passionately held dream to devise smart, brain/machine interfaces that will redress serious health problems, as well as potentially to augment our native capabilities.”
In recent weeks, Nicolelis also reported on a rat fitted rat with a sensor that allowed the animal to detect invisible infrared light and he has plans for a brain-controlled robotic exoskeleton that a handicapped child would demonstrate either in the 2014 World Cup or the 2016 Olympics, both in his home country of Brazil. So what’s the frequency, Miguel? Stay tuned for more BrainNet.
Image Source: Katie Zhuang, Nicolelis Lab, Duke University
Get 6 bi-monthly digital issues
+ 1yr of archive access for just $9.99