Rats Communicate Through Brain Chips
Pairs of rats can communicate through brain chips and collaborate to perform a task, report researchers in today’s Scientific Reports. Brain activity recorded in one rat was translated into a pattern of electrical pulses that were then transmitted to another rat that had been trained to push a particular lever in response to one of two patterns of electrical stimulation in its brain. The rats also worked together, say the researchers. If the second rat chose the wrong lever, then the first rat would change its brain function and behavior in the next trial so that the receiving rodent was more likely to get it right, claim the scientists.
The research was led by Miguel Nicolelis, a neuroscientist at Duke University Medical Center, who has previously described a brain-computer interface through which a monkey could control a walking robot (see “The Power of Thought”) and another setup in which a virtual sense of touch was fed into a monkey’s brain through an electrical stimulating array (see “Giving Prosthetics a Sense of Touch”). A handful of labs have been making impressive progress in reading and writing to the brain in recent years with the aim of helping paralyzed people regain mobility via thought-controlled robotics. Last year, two research teams reported that quadriplegic patients could use brain implants to control robot limbs (see “Brain Chip Helps Quadriplegics Move Robotic Arms with Their Thoughts” and “Patient Shows New Dexterity with a Mind-Controlled Robot Arm”).
But today’s study, says Nicolelis, was not about improving brain-computer interface technology for patients but rather exploring new frontiers. “We observed the emergence of physiological properties that we could not predict before we did this,” he says, pointing to what he calls collaboration between the two animals’ brains.
In the experiment, Nicolelis and his team trained a rat to choose between a right-side or left-side lever to push depending on which of two LEDs lit up. If the rat pushed the correct lever, it got a rewarding sip of water. The researchers recorded the electrical activity of the rat’s motor cortex, the region of the brain that controls movements, and translated the activity involved in pushing the right-side lever into many pulses and pushing the left-side lever into fewer pulses. These pulses were then sent to the implant in the brain of another rat in a separate chamber. That rat had been trained to respond to pulse patterns in a similar way—more pulses meant push the right-side lever.
With no cue from the LEDs in its cage, the second rat was able to choose the correct lever 64 percent of the time, at which point both rats would get a water reward (the information-sending rat would thus get two; the information receiving rat would get only one). When the second rat got it wrong, the first rat noticed, says Nicolelis, because it did not get a second reward. So in the next trial, the first rat would respond more quickly to the LED cue and produce a greater amount of task-related neuron firing compared to background brain noise, he says, which made the second rat more likely to choose the correct lever. This is what Nicolelis refers to as collaboration.
The researchers also demonstrated the brain-to-brain communication with whisker stimulation in the first rat. Like a cat, rats use their whiskers to determine how wide an opening is, and the rodents can be trained to turn their head to the left or right depending on whether a hole in their cage is narrow. Similar to the first experiment, the brain activity of the first rat was translated into a particular pattern of pulses sent to the second rat, which had been trained to poke its head left in response to electrical pulses, and right in the absence of pulses. With these tests, the second rat chose the correct side about 62 percent of the time.
With the whisker test, the team demonstrated that the rats need not be in the same building—or even on the same continent—to collaborate. A rat in Brazil at the Edmond and Lily Safra International Institute of Neuroscience of Natal sent brain signals to a rat on the Duke campus in Durham, North Carolina.
However, the binary decisions made in the rat tests are not up-to-speed with what brain-computer interfaces can do these days, wrote University of Pittsburgh’s Andrew Schwartz, a pioneer in patient brain-computer interfaces, in an e-mail to MIT Technology Review. “It may sound like ‘mental telepathy’ and therefore seem exciting, but when looked at more carefully, it is very simplistic,” he wrote. “As a communication channel, you could think of a locked-in patient trying to communicate by blinking, where a blink means yes and no blink mean no. This kind of information could be conveyed by recording from a single neuron in one rat and buzzing electrical current in the receiver rat. If the rat feels the buzz, it means yes, no buzz means no.”
But Nicolelis sees this demonstration as the beginning of a new line of research that could lead to a new form of computing. He says his lab is working on “swarms” of rats that could share motor and sensory information via brain-to-brain interfaces. “If you put brains together, you could create a more powerful non-Turing machine, an organic computer that computes by experience, by heuristic,” he says. “That could be a very interesting architecture to explore.”
Geoffrey Hinton tells us why he’s now scared of the tech he helped build
“I have suddenly switched my views on whether these things are going to be more intelligent than us.”
Meet the people who use Notion to plan their whole lives
The workplace tool’s appeal extends far beyond organizing work projects. Many users find it’s just as useful for managing their free time.
Learning to code isn’t enough
Historically, learn-to-code efforts have provided opportunities for the few, but new efforts are aiming to be inclusive.
Deep learning pioneer Geoffrey Hinton has quit Google
Hinton will be speaking at EmTech Digital on Wednesday.
Get the latest updates from
MIT Technology Review
Discover special offers, top stories, upcoming events, and more.