Click the link below the picture
.
For people with near-total paralysis, the ability to communicate easily in real time is a challenge. For years, scientists have been working to remedy that by developing devices that can decode brain signals and translate them into computer cursor movements or text.
These devices are a type of brain-computer interface, or BCI, and they consist of electrode chips that are implanted inside the brain to listen to and decode the electrical whispers of neurons. In the past, BCIs allowed people to type using a virtual keyboard, but the speed was frustratingly slow. Now, however, a team of scientists report that their BCI keyboard helped two people with paralysis type at speeds of up to 22 words per minute—nearly as fast as the average person can text using a smartphone. The findings were published today in Nature Neuroscience.
“This is an important technical advance that brings brain-computer typing much closer to practical communication speeds for people with paralysis,” says Edward Chang, a professor of neurological surgery at the University of California, San Francisco, who was not involved in the study.
“At about 22 words per minute, this is among the fastest motor-cortex typing BCIs yet and dramatically faster than most earlier neural spellers,” says Chang, who has worked on another speech-decoding BCI.
BCI technology has advanced significantly since its genesis in the 1960s, when researchers began using single electrodes implanted in the brains of monkeys to record their neural activity. In 2006, a consortium of researchers called BrainGate reported that a BCI allowed people with paralysis to control a computer cursor and operate a prosthetic hand. In recent years, the BrainGate BCI was used to control a virtual keyboard with a cursor and to decode letters from handwriting areas of the brain. Other groups’ BCIs have decoded words or short phrases directly from speech-related brain regions, too.
Previous versions of these brain-typing systems required participants to control a cursor on a screen and individually select letters, “which is far slower than being able to access any key at any time using your fingers,” says lead study author Justin Jude, a postdoctoral researcher at BrainGate, based at Brown University, and an appointed research fellow at Massachusetts General Hospital and Harvard Medical School.
In the new paper, Jude and his colleagues trained their BCI using artificial intelligence to recognize intended hand or finger movements from a part of the brain’s movement area, called the precentral gyrus, as participants tried to move their paralyzed hands or fingers. The AI model predicted the letters on a QWERTY keyboard that the movements most likely corresponded to. They tested their system in two participants: one person with amyotrophic lateral sclerosis, a progressive neurological disease that causes paralysis, and one with a spinal cord injury that left them paralyzed but still able to speak.
Using the device, the latter participant was able to type at 110 characters or 22 words per minute, with a word error rate of 1.6 percent. The other participant’s typing was slower but still impressive for someone who lacked the ability to speak. By comparison, several years ago, the BrainGate handwriting BCI achieved speeds of 90 characters per minute (about 18 words per minute). Another previous BCI that was implanted in a speech-related brain region by Chang and his colleagues was used to achieve a typing speed of 78 words per minute, but the median word error rate was far higher—25 percent.
“One of the things that we talk about a lot is the speed of communication. The reason we do that is not just to have a faster system than someone else,” says Daniel Rubin, a critical care neurologist at Massachusetts General Hospital and an assistant professor of neurology at Harvard Medical School, who was a co-author of the new study.
People who have lost the ability to speak and to use their hands might be able to employ an eye-tracking system to type, but this method is slow. “Communication speed matters, because being part of a conversation matters,” Rubin says.
The researchers say the QWERTY keyboard system is more successful than the version that decoded mental handwriting. It remains to be seen, however, whether decoding from brain regions that control finger movement or from speech-related regions is a better strategy overall, Chang says. Signals in the brain’s motor cortex are easier to decode, but those in speech-related areas might be faster and more direct.
The technology is not ready for widespread use yet; the study was small, and the device requires brain surgery, which carries risks. “The biggest limitations are the small number of participants and the need for invasive intracortical implants,” Chang says.
Another limitation is the need to calibrate the BCI each time before it can used. “It’s almost like a musical instrument, and you have to tune it each day,” Rubin says. Having an instrument that can tune itself is a big goal for the field, he says.
Several companies are developing commercial BCIs, primarily for use in people who are paralyzed. Perhaps the most hyped has been Elon Musk’s Neuralink, but there are others, such as Paradromics and Synchron. (Some of the study authors consult for these companies and receive research funding from them.)
China recently approved the first invasive BCI for use in people with a form of partial paralysis. No such devices have been approved by the Food and Drug Administration for use by people with paralysis in the U.S.
.
A BrainGate participant types using a brain implant. BrainGate Consortium
.
.
Click the link below for the complete article:
.
__________________________________________
Leave a Reply