This mechanical arm controlled by the brain can twist, grasp and feel

[ad_1]

The brain is two-way: it absorbs information and at the same time sends signals to other parts of the body, telling the brain to take action.Even an action that looks as simple as grabbing a cup will call your brain and command your hand muscles and listen To the nerves of the fingers.

Because Copeland’s brain was not injured in the accident, theoretically, it can still manage input and output dialog boxes. However, most of the electrical information sent by the nerves in his body did not reach the brain. When the Pittsburgh team recruited him to their research room, they wanted to devise a solution. They believe that the brain of a paralyzed person can either stimulate the robotic arm or be stimulated by electrical signals from the robotic arm, and ultimately interpret this stimulation as the sensation of being touched by their own hands. The challenge is to make everything natural. When Copeland wants to twist, the mechanical wrist should also twist. When he intends to grasp, the hand should be closed; when the robot’s little finger touches a hard object, Copeland should feel it with his little finger.

In the four microelectrode arrays implanted in the Copeland brain, two grids read the motion intention from his motor cortex to command the robotic arm, and two grids stimulate his sensory system. From the beginning, the research team knew that they could use BCI to create tactile sensations for Copeland by simply delivering current to these electrodes, without the need for physical touch or robotic manipulation.

To build the system, the researchers took advantage of the fact that Copeland’s right thumb, index and middle fingers retain a certain sensation. When the researcher was sitting on the magnetic brain scanner, he rubbed a Q tip there, and they discovered which specific contours of the brain corresponded to those fingers. Then, the researchers decoded his movement intention by recording the brain activity of a single electrode and imagining a specific movement. He felt it when they connected electricity to specific electrodes in his sensory system. To him, this feeling seemed to come from the roots of his fingers, near the top of his right palm. It felt like natural pressure or warmth, or strange tingling-but he had never experienced any pain. “Actually, I kept staring at my hand, like,’Man, that really feels like someone can poke there,'” Copland said.

Once they have determined that Copeland can experience these sensations, and the researchers know which brain regions to stimulate to produce sensations in different parts of his hand, the next step is to get Copeland to get used to controlling the robotic arm. He and the research team set up a training room in the laboratory and hung up posters of Pac-Man and cat memes. Three days a week, the researcher hooked electrode connectors from his scalp to a set of cables and computers, and then when he grabbed the blocks and balls and moved them from left to right, they would time him. In the past few years, he has performed very well.He even Demonstrated the system The then President Barack Obama (Barack Obama).

But then, Collinger said: “He is in a stable state at a high level of performance.” A non-paralyzed person takes about five seconds to complete the task of moving objects. Copeland can sometimes complete the operation in six seconds, but his median time is about 20.

In order to get him out of trouble, it is time to try to provide him with real-time touch feedback from the robot arm.

Human fingers feel the pressure, and the electrical signals generated will slide along the linear axons from the hand to the brain. The team reflected the sequence by placing sensors on the fingertips of the robot. But objects don’t always touch the fingertips, so a more reliable signal must be sent from other places: the torque sensor located at the bottom of the robotic finger.

[ad_2]

Source link