The ability to control a robot with your thoughts may sound like something from a Spielberg movie. But it may soon become reality for people with paralyzing diseases.
Running robots with brain power. I'm Bob Hirshon and this is Science Update.
Right now, people with paralyzing injuries or diseases can only imagine what it's like to move. But in the future, imagining may be enough to control robotic limbs that can assist them.
That's the goal of Duke University neurobiologist Miguel Nicolelis. By linking brain implants to a computer, he and his colleagues have already made it possible for monkeys to control a robotic arm, just by thinking about it. Recently, they captured the brain activity of human patients, who were asked to play a simple video game.
We are basically recording the pattern that is produced in the brain of these patients as they are preparing to make these movements.
Then they taught a robotic arm to respond to the brain signals, much like a human arm would. After more animal studies, the next step is to put it all together in paralyzed patients, using permanent brain implants.
And in fact we already have a wireless interface that will allow us to broadcast the raw signals coming from the brain, to a microchip, that is going to do all [of] the signal processing required to convert these signals into motor commands that a robot can read.
It's hoped that the first human trials will begin by 2005. I'm Bob Hirshon for AAAS, the Science Society.
Making Sense of the Research
Many people who are paralyzed still have some movement: in their upper body, or in their face and eyes, for example. But some people are so completely paralyzed that they can't even speak, much less move. Giving them the power to control something—anything—on their own would give them a new kind of life. Even for patients who are only partially paralyzed, thought-controlled robotics could offer a new kind of freedom.
Here's how Nicolelis and his colleagues have built their system. First, they studied the brain patterns of monkeys as they played a simple video game. Whenever the monkey moved its arm to the left, the scientists recorded the bursts of brain activity that went along with the movement. After enough trials, they had a reasonably good idea of what the command "move your arm to the left" looks like. They did the same for a number of other simple movements.
Next, they taught a robot arm to respond to the patterns of brain activity in a way that mimicked real life. So, for example, they designed the robot arm to move to the left whenever it received "move arm left" patterns of brainwaves. Then they hooked up the monkey's brain to the robot arm, through a computer, and—presto!—the monkey controlled the arm with its thoughts.
The nice thing about this design is that the scientists didn't have to understand exactly how the brain was creating the movements. All they needed to know was what the brain patterns looked like, not what they meant or how they worked. It's kind of like listening to somebody give commands to a dog in a foreign language. Suppose every time the owner said "Blorg smerkle!," the dog sat down. If you said, "Blorg smerkle!" to the dog, the dog would probably sit too. You don't have to know what "Blorg smerkle!" actually means. And if you watched the owner long enough, you could control the dog very well without ever having to learn or understand the owner's language.
The monkey experiments were done when the monkey's skulls were open and brains exposed, which you can't do to humans just for a study like this. But Nicolelis and his colleagues got a chance to work on humans who were having brain surgery anyway. Because their time with the patients was very limited, the researchers could only record the subjects' brain activity, and then teach the robot arm to respond to different brain patterns in the lab. In other words, humans have not controlled the arm with their thoughts yet. But there's no reason why it shouldn't work just as well.
Eventually, the researchers hope to fit paralyzed patients with permanent brain implants to use with the device. Once they are installed, they could transmit information through wireless networking, just like you would use to surf the Internet from a laptop or a cell phone. The result would take remote control to a whole new level.
Now try and answer these questions:
- How are Nicolelis and his colleagues creating their thought-controlled robots?
- What could a thought-controlled robot made in this way potentially do? What would its limitations be?
- Suppose it turned out that when people lose the ability to move, their brains process thinking about movement in different ways. How would this affect this technology?
- If you were studying the first human clinical trial of this technology, what questions would you want to answer?
Miguel Nicolelis' Lab has a comprehensive website with information about this project and many similar lines of research.
Monkey brain operates machine, from BBC News, describes Nicolelis' first successes with monkeys in 2000.