Abraham lincoln

Descargar 2,95 Mb.
Fecha de conversión05.08.2017
Tamaño2,95 Mb.
1   2   3   4   5   6   7   8   9   10   11

Of course, spending all those resources to get comprehensible signals into and out of a brain means buying a central assumption: that the brain is the center of control for the limbs. But what if it isn’t? It takes 300 milliseconds, almost a third of a second, for the human arm to send a message to the conscious brain and for the brain to respond. If that was the only way to control the hand, we’d never manage to balance trays or hang up clothes. Those activities require a degree of fine-motion control that seems to outpace the speed of signals going to and from the brain.

That’s what makes some researchers think that the brain has learned to delegate fast-response tasks to the spinal cord. It’s closer to the arm and can respond up to 10 times faster—in just 30 milliseconds. “The moment-to-moment timing of the hand’s muscle contractions is dependent on sensory feedback that is never going to the brain,” says Loeb, the USC biomedical engineer. “It’s being handled locally.” In other words, the spinal cord isn’t just a dumb trunk line. It’s a coprocessor.

Unfortunately, tapping into the spinal cord is even tougher than tapping into the brain. “It’s really hard,” Donoghue says with a rueful laugh. “The spinal cord moves around a lot, so the mechanical problem is huge. Sticking an electrode in there that stays in place is a big challenge.”

Loeb has created a rudimentary “virtual spinal cord,” a software-based substitute for the fine motor control provided by the spine. Theoretically, a technology like this could decode the brain’s high-level intentions and issue supplementary fine motor commands to a prosthetic limb. Easier said than done; inferring a brain’s intentions is the central challenge of brain-machine interfaces. “If the bionic hand has this sort of mind of its own,” Loeb says, “what kind of command signals do we need to get from the brain to control this semiautonomous beast?”

Still, researchers in the field remain stubbornly hopeful. At Stanford, Shenoy is now embedding wireless neural interfaces in monkeys and then recording their movements with digital cameras. He’s trying to correlate the neural data with wireframe graphics of the monkeys in motion. “By measuring how the brain controls the arm under more natural, real-world conditions,” Shenoy says, “we’ll at least learn how the neural activity relates to the arm in those other situations.” But the data sets will be enormous. He thinks algorithms for “dimensionality reduction”—reducing to just a few critical features what the computer has to interpret—might take care of the problem. It’s pretty abstract math, but it’s a way to teach a computer how to distinguish, let’s say, among hundreds of breeds of dogs by pulling out just a few variables, like coat color or ear shape. The problem is, dimensionality reduction works best with a ton of raw data—input from hundreds of thousands of neurons would be ideal.

It’ll be a long time before anyone can monitor that many. The number of neurons we can record at one time has doubled just about every seven and a half years since 1959. At this rate, it should be possible to monitor 1,000 neurons by 2026. And researchers will be able to track all 100 billion neurons in a human brain in a mere 220 years.

But what about now? Todd Kuiken ticks off a list of things that patients can do today with an arm like Glen Lehman’s: take out garbage, put on socks, open a jar, pick up a hat. Optimism seems reasonable in the long run. Arms and spinal cords and brains are complex, but they’re not magic.

Last year, Darpa appeared to declare victory over the problem. The agency changed the language on its website to say that its researchers had “delivered a prosthetic arm for clinical trials” that did everything they promised it would in 2006, including neural control. There is no such arm; a Darpa public affairs officer says that the wording of the recent website post was incorrect, but the agency’s head of prosthetic research wouldn’t answer further questions.

Back home in Pennsylvania, Glen Lehman straps on an old-school cable-controlled mechanical arm with a split hook for gardening or working around the house. “It can take more abuse,” he says, “and I can make some of the repairs on my own.” And every few months Lehman heads back to the Rehabilitation Institute of Chicago to practice with his experimental limb and rewired nerves. This version isn’t anywhere near perfect, but there’s always next time. And the time after that.

  1. Pruebas con el mindwave

Además de esto, quise comprobar esto a través de varios otros experimentos parecidos, y midiendo sus niveles de concentración con el “mindwave”.

  1. Rodrigo Fribourg

  2. Gonzalo Ueda

En el primer experimento quise comprobarlo viendo que tanto se concentraban al hacer rebotar una pelota de Ping Pong en la raqueta, si bien es cierto que ninguno de los 2 son muy buenos en Ping Pong, Rodrigo Fribourg es un poco mejor que Gonzalo Ueda.

1   2   3   4   5   6   7   8   9   10   11

La base de datos está protegida por derechos de autor ©absta.info 2016
enviar mensaje

    Página principal