No fictional plot: paralyzed man communicates with people through imaginary handwriting

“Writing text by thinking” is no longer a science fiction plot! The journal “Nature” (Nature) reported on the 12th, an experimental device that turns thoughts into text, allowing a completely paralyzed man to imagine himself handwriting text on paper and quickly build words on a computer screen with an accuracy rate of up to 95 percent.

Jaimie Henderson, a member of the research team and a neurosurgeon at Stanford University, said the device is so useful for people who can’t move or speak that “we can foresee a future where people with spinal cord injuries can use e-mail and even computer programmers can go back to work even if they are paralyzed. “

And Krishna Shenoy, a researcher at Stanford University and the Howard Hughes Medical Institute who co-directed the research team with Henderson, said, “Amazingly, we found that he could type 90 characters per minute. “

Both Henderson and Shenoy are highly interested in “commercializing” this experimental method of “decoding brain signals.

John Ngai, head of the BRAIN Initiative at the National Institutes of Health, which sponsored the research, said the idea of decoding the brain’s handwriting activity was “just brilliant. “; “it’s just one of those subjects in a laboratory setting; but it’s an excellent demonstration of the principle.”

Henderson noted that the man who agreed to participate in the test slipped and fell while taking out the trash and became quadriplegic. A few years ago, the man agreed to participate in an experimental study called BrainGate2, an experimental system that implants electrodes near the blocks of the brain that control movement, with the goal of allowing paralyzed people to control computers and other devices by thought alone. In the past, participants have learned to control computer cursors or robotic arms by imagining they are doing it.

This time, Henderson, Shenoy and the team of scientists had the participant imagine that he was writing a personal letter by hand while the computer monitored his brain’s electrode activity; the computer eventually learned to decode all the letters with specific patterns of activity associated with multiple symbols, identify the letters the participant wrote, and then put them on the screen to spell out words and sentences.

Past experiments have tried to have participants imagine themselves clicking on letters on the computer screen and found that “imagining handwriting” was much faster than “imagining clicking on the screen. And, because the new experimental system relied on already familiar ideas, the subjects were “on board” almost immediately. Henderson said, “He was very excited when he could respond in writing to the questions we asked him.”

John Ni said the team’s success in decoding imaginary handwritten text is the latest advance in connecting computers to the human brain, “I heard about this concept more than a decade ago, and at the time I thought it was like science fiction. But about five years later, experiments proved it wasn’t fantasy! Now we’re seeing great progress, and it’s really exciting.”