Scientists in the United States have created a sensor that can be implanted in the brain that allows you to think about letters and see them appear written on a computer screen. This technology could give paralyzed people the ability to communicate in writing without using their hands and faster than other interfaces, said Krishna Shinui, a researcher at Stanford University, one of the authors of the study published Wednesday in the scientific journal. nature.
For this study, the researchers decoded Nerve activity that occurs in the brain When trying to write letters by hand and implanting a sensor in the brain of a paralyzed person with a cervical spine injury.
Paul Scoma
Then they used an algorithm to determine the letters when the volunteer tried to write them, translating the neural activity in real time into the characters that appeared on the screen. Since these characters are handwritten, they average 90 characters per minute, more than double what was achieved with them. An interface between the brain and the computer in previous experiments.
Sums up researcher Jose Carmina, UC Berkeley researcher, who sees The study in nature “Significant progress in this area.” “This system utilizes both fertile neural activity recorded by electrodes between the cortex and the power of language models that, when applied to decoded characters, can generate text quickly and honestly,” said lead researcher Frank Willett.
Willett et al / Nature 2021
The study volunteer was a 65-year-old man who was paralyzed from the neck down. In this experiment, two aspirin-sized electrodes were placed on the part of the associated brain responsible for the movement of his right arm and right hand.
Through the use of signals that Detected sensors in individual neurons When a person imagined writing, the algorithm recognized the patterns that his brain produces when it thinks about each letter. Using this system, the man was able to transcribe phrases and answer questions at the same rate that a person could do when typing on a cell phone. The team suggests that the speed of the system is due to the specific activity that each letter triggers in the brain, which allows the algorithm to distinguish it.
The study was conducted inside Braingate collaborative projectWhich brings together researchers from Brown, Harvard, Stanford and Case Western Reserve universities, as well as Massachusetts Hospital and Providence Virginia Medical Center in Rhode Island.
“Friendly zombie fanatic. Analyst. Coffee buff. Professional music specialist. Communicator.”
