Neural Implant Lets Paralyzed Person Type By Imagining Writing (arstechnica.com) 25
An anonymous reader quotes a report from Ars Technica: This week, the academic community provided a rather impressive example of the promise of neural implants. Using an implant, a paralyzed individual managed to type out roughly 90 characters per minute simply by imagining that he was writing those characters out by hand. Somewhere in our writing thought process, we form the intention of using a specific character, and using an implant to track this intention could potentially work. Unfortunately, the process is not especially well understood. Downstream of that intention, a decision is transmitted to the motor cortex, where it's translated into actions. Again, there's an intent stage, where the motor cortex determines it will form the letter (by typing or writing, for example), which is then translated into the specific muscle motions required to perform the action. These processes are much better understood, and they're what the research team targeted for their new work. Specifically, the researchers placed two implants in the premotor cortex of a paralyzed person. This area is thought to be involved in forming the intentions to perform movements. Catching these intentions is much more likely to produce a clear signal than catching the movements themselves, which are likely to be complex (any movement involves multiple muscles) and depend on context (where your hand is relative to the page you're writing on, etc.). With the implants in the right place, the researchers asked the participant to imagine writing letters on a page and recorded the neural activity as he did so.
Altogether, there were roughly 200 electrodes in the participant's premotor cortex. Not all of them were informative for letter-writing. But for those that were, the authors performed a principal component analysis, which identified the features of the neural recordings that differed the most when various letters were imagined. Converting these recordings into a two-dimensional plot, it was obvious that the activity seen when writing a single character always clustered together. And physically similar characters -- p and b, for example, or h, n, and r -- formed clusters near each other. (The researchers also asked the participant to do punctuation marks like a comma and question mark and used a > to indicate a space and a tilde for a period.) Overall, the researchers found they could decipher the appropriate character with an accuracy of a bit over 94 percent, but the system required a relatively slow analysis after the neural data was recorded. To get things working in real time, the researchers trained a recurrent neural network to estimate the probability of a signal corresponding to each letter.
Despite working with a relatively small amount of data (only 242 sentences' worth of characters), the system worked remarkably well. The lag between the thought and a character appearing on screen was only about half a second, and the participant was able to produce about 90 characters per minute, easily topping the previous record for implant-driven typing, which was about 25 characters per minute. The raw error rate was only about 5 percent, and applying a system like a typing autocorrect could drop the error rate down to only 1 percent. The tests were all done with prepared sentences. Once the system was validated, however, the researchers asked the participant to type out free-form answers to questions. Here, the speed went down a bit (to 75 characters a minute) and errors went up to 2 percent after autocorrection, but the system still worked. The findings have been published in the journal Nature.
Altogether, there were roughly 200 electrodes in the participant's premotor cortex. Not all of them were informative for letter-writing. But for those that were, the authors performed a principal component analysis, which identified the features of the neural recordings that differed the most when various letters were imagined. Converting these recordings into a two-dimensional plot, it was obvious that the activity seen when writing a single character always clustered together. And physically similar characters -- p and b, for example, or h, n, and r -- formed clusters near each other. (The researchers also asked the participant to do punctuation marks like a comma and question mark and used a > to indicate a space and a tilde for a period.) Overall, the researchers found they could decipher the appropriate character with an accuracy of a bit over 94 percent, but the system required a relatively slow analysis after the neural data was recorded. To get things working in real time, the researchers trained a recurrent neural network to estimate the probability of a signal corresponding to each letter.
Despite working with a relatively small amount of data (only 242 sentences' worth of characters), the system worked remarkably well. The lag between the thought and a character appearing on screen was only about half a second, and the participant was able to produce about 90 characters per minute, easily topping the previous record for implant-driven typing, which was about 25 characters per minute. The raw error rate was only about 5 percent, and applying a system like a typing autocorrect could drop the error rate down to only 1 percent. The tests were all done with prepared sentences. Once the system was validated, however, the researchers asked the participant to type out free-form answers to questions. Here, the speed went down a bit (to 75 characters a minute) and errors went up to 2 percent after autocorrection, but the system still worked. The findings have been published in the journal Nature.
Neat But Scary Where This Can Go (Score:5, Insightful)
The potential for this is pretty incredible both for good AND bad.
Re:Neat But Scary Where This Can Go (Score:4, Funny)
We're detecting some thoughtcrime from your cortex, citizen. Please report to room 101 in the basement of the Ministry of Love.
Re: Neat But Scary Where This Can Go (Score:2)
Re: Neat But Scary Where This Can Go (Score:2)
I believe your post is a dupe, originally from fifty years ago =)
Re: (Score:1)
If one thinks in Russian the launch potential is amazing.
Re: (Score:2)
Re: (Score:2)
The potential for this is pretty incredible both for good AND bad.
Not so fast, because this is still an implant technology. But how soon will it become a helmet technology?
Re: (Score:1)
This guy could be our ASCII art lameness filter. Imagine what happens when he sees goatse! Then what gets printed on the screen?
Error rate (Score:4, Funny)
Neural Implant Lets Paralyzed Person Type By Imaging Writing
By imaging writing? Seems slashdot editors could use this neural implant to improve their error rate.
Re: (Score:2)
It turns out that imaging writing is only 90% accurate.
Imagine if there was some kind of system that allowed colleagues to check each other's work before releasing it to the public. That's a bajillion dollar idea right there waiting for someone to implement.
Neuralink (Score:2)
Re: (Score:2)
But I guess you think this is okay because Elon is not involved.
We think it's ok because paralyzed people went from mind-typing 25 characters per minute to mind-typing 90 characters per minute. Unless and until somebody figures out nerve regeneration, stuff like this is a fuckton better than nothing, regardless of who does it.
No one who isn't paralyzed should be volunteering to have their brain stuffed full of electrodes just yet. Even hair thin electrodes do plenty of damage. But for paralyzed people, it's a.... *puts on sunglasses* no brainer.
Re: Neuralink (Score:3)
Krishna Shenoy is a cofounder of Neuralink. This work is from Krishna Shenoyâ(TM)s group at Stanford.
Re: (Score:2)
Re: (Score:2)
With Neuralink, do they even know how to remove the implants? Not many people would line up for a neural implant if they can't remove them yet.
Medicines, vaccines, surgical procedures, genetic engineering, cancer research, toxicology tests -- it's all done on animals first. There are strict guidelines against doing testing on humans until you have a clinical trial and adequate safety protocols. The more cutting edge, the more animal testing that is involved.
Also, for many things, testing on humans is unr
Re: (Score:2)
"Imaging"? (Score:2)
Please fix the headline! They aren't "imaging writing" they are "imagining writing"
Re: "Imaging"? (Score:3)
A whole LOAD of things need to be fixed on /. and it ain't just grammar..
Better not get... (Score:1)
How long does the implant last? (Score:3)
I did not find any data about this one, but some other similar experiments had to remove the sensors after a few months to avoid problems. Long-term bio-compatible sensors are still a quite unsolved problem and sensors connected to nerves or the brain are especially problematic.
But when will it support (Score:2)
Unicode.