Sending Messages With Your Brain Via EEG

An anonymous reader writes "From a University of Wisconsin-Madison announcement: 'In early April, Adam Wilson posted a status update on the social networking Web site Twitter — just by thinking about it. Just 23 characters long, his message, 'using EEG to send tweet,' demonstrates a natural, manageable way in which "locked-in" patients can couple brain-computer interface technologies with modern communication tools. A University of Wisconsin-Madison biomedical engineering doctoral student, Wilson is among a growing group of researchers worldwide who aim to perfect a communication system for users whose bodies do not work, but whose brains function normally.' A brief rundown of the system: Users focus on a monitor displaying a keyboard; the interface measures electrical impulses in the brain to print the chosen letters one by one. Wilson compares the learning curve to texting, calling it 'kind of a slow process at first.' But even practice doesn't bring it quite up to texting speed: 'I've seen people do up to eight characters per minute,' says Wilson. See video of the system in action."

Comments

TCMP?

[explosivejared]

Is this anything like TCMP?

This may be the first time

[Anonymous Coward]

That a brain was involved in the process of Tweeting.

8 characters a minute is excellent.

[Finallyjoined!!!]

Think how much more Stephen Hawking could gives us with this device.

I know he's in the hossie at the moment and I hope he recovers fully, enough to try this device. :-)

Send one to him. Now!

Re:8 characters a minute is excellent.

[Anonymous Coward]

"hossie"? What are you, 6 years old?

Re:8 characters a minute is excellent.

[zappepcs]

If they are not already in contact with him about this, they probably don't deserve to see any profit from it. This is the first thing I thought of since the stories are right next to each other.

I'm also interested to know if they can improve this to work even when people can't see the keyboard etc.

Re:8 characters a minute is excellent.

[Finallyjoined!!!]

Why would they call him a cigarette??

How odd.

Re:8 characters a minute is excellent.

[Penguinshit]

Professor Hawking uses a more sophisticated system with word-prediction and a micro-switch activated by a slight motion of his shoulder. he can do much better than 8 cpm. I use a similar system but use eye gaze on a virtual keyboard rather than a sectoring keyboard.

Perhaps he's more accustomed to the sectoring keyboard or no longer has the ocular control for the eye gaze system.

Re:8 characters a minute is excellent.

[jd]

The eyes probably couldn't be steered accurately enough. His muscular control was a mess when I saw him in person in the late 1980s, and it won't have improved since.

On the other hand, if they tune into the neurons that control his arm, they may be able to anticipate what he is going to type. That might help accelerate things for him. It's a bit much to be able to decode the language centres sufficiently to record thoughts directly, but it will eventually get to that point.

Once it is possible to decode his thoughts directly, he would be able to communicate as fast as he can think. Which means that it'll be a babble because he thinks far too fast. On the other hand, it will help him to turn out papers at a fantastic speed.

The website for the lab...

[Saint Aardvark]

...can be found here:

http://nitrolab.engr.wisc.edu/

Re:Not as quick as texting...yet

[arth1]

I would think that this would be an excellent application for Dasher, which should bring the speed way up above 8 cpm.

Eye tracking?

[TinBromide]

So, when the letter being focused on flashes, the EEG picks it up and figures out which row and column are desired...

So it wouldn't work very well for the blind and its not pulling the letters out of the brain, its just a more sophisticated eye tracking device, similar to the goggles in apache helicopters? Why not just fit patients with those for a faster input method?

Re:Eye tracking?

[arth1]

Now you're not only assuming that the user can move his eyes, but also blink at will.
The whole point is to bring communications to the very worst cases, who currently have no methods of communication because they can't control their body at all.

Re:Eye tracking?

[Chris Burke]

So it wouldn't work very well for the blind and its not pulling the letters out of the brain, its just a more sophisticated eye tracking device, similar to the goggles in apache helicopters? Why not just fit patients with those for a faster input method?

Because Apache helicopters are prohibitively expensive even for patients with the best insurance, aside from being illegal for civilians to own. Duh.

Re:Eye tracking?

[RDW]

'Because Apache helicopters are prohibitively expensive even for patients with the best insurance, aside from being illegal for civilians to own.'

It would probably be much cheaper to pick up a surplus thought-guided control system from the Soviet Mig-31 project on ebay. The only downside (and this is very important) is that you must think in Russian. You can't think in English and transpose it - you must think in Russian.

f i r s t p o s t

[Anonymous Coward]

d a m n  t h i s  t h i n g  i s  s l o w

Hell of a fruedian slip

[Cyberwasteland]

As this technology gets better isn't there going to be a big chance for really bad Fruedian slips? XD

I see a way to speed it up.

[Ungrounded Lightning]

Instead of flickering one row or column at a time, flicker ALL the letters simultaneously in different patterns. The brainwave trace should follow the one you're watching and the wait for it to be identified and confirmed will be much shorter.

= = = =

How is this better than eye tracking?

Optimization

[Rival]

FTA:

"The interface consists, essentially, of a keyboard displayed on a computer screen. "The way this works is that all the letters come up, and each one of them flashes individually," says Williams. "And what your brain does is, if you're looking at the 'R' on the screen and all the other letters are flashing, nothing happens. But when the 'R' flashes, your brain says, 'Hey, wait a minute. Something's different about what I was just paying attention to.' And you see a momentary change in brain activity."

Their "cognitive click from flash recognition" interface sounds an awful lot like the retrace timing system used for the http://en.wikipedia.org/wiki/NES_Zapper.

I'm curious what kind of language optimization has been added, if any. Do they use predictive text of some sort?

Also, it seems a waste to limit the input to a display of a static keyboard (other than ease of use for people who know where to look for certain letters.) Why not have a dynamic interface, something alongs the lines of http://en.wikipedia.org/wiki/Dasher/?

Re:But you need a screen?

[arth1]

Because the user might not be able to move his eyes.
The idea is that no focusing is required, just thought.

Re:Not as quick as texting...yet

[arth1]

Especially if your name is Etaoin Shrdlu.

Obligatory...

[jeffmeden]

Kif Kroker: One beep for yes, two beeps for no.
[Fry beeps once] ... [Fry beeps twice]
Captain Zapp Brannigan: Double yes. Guilty.

Re:Not as quick as texting...yet

[johny42]

I wonder how fast he would be if he imagined a Dvorak keyboard.

P300 Explanation and Anecdotal Info

[Beefmancer]

This system has been around for a while; I've seen it demonstrated live twice, and it didn't work at all either time. In my opinion, even in best conditions (bald patient, shit-tons of electrodes, professional setup, well-trained subject) it doesn't work well enough to fuel science-fiction fantasies, and probably never well. For locked-in patients, who can do nothing but move their eyes, though, it's an awesome technology. They made a movie recently about such a patient who spent years using it to write a novel: http://www.imdb.com/title/tt0401383/.

Here's how it works - metal electrodes on the brain (EEG) pick up an analog signal, and *any* stimulus which is particularly salient to a subject creates a spike in the signal 300ms after that stimulus appears (this spike is called the P300, there's a good wiki article on it). If you have a dude staring at a grid of letters, you can tell which one he's looking at by hunting for the big spikes 300 ms after the right letter flashes. The only problem is the signal-to-noise ratio, which is notoriously terrible in EEG, though of course there are people out there working on improving it.

Re:Not as quick as texting...yet

[treeves]

Good one. I always employ that sequence (I learned originally in Hoftstadter's Godel, Escher, Bach) when I play Hangman. I never lose.

Re:Not as quick as texting...yet

[Penguinshit]

I used to use Dasher before ALS took my hands completely. It flies by way too fast and the letter choice way too "random" to be useful in the current application.

Perhaps if the sensing algorithm gets an order of magnitude faster and more precise, but in the average lifetime of a person diagnosed today with ALS or locked-in syndrome it will not happen.

Re:Human Computer Interaction

[jd]

Similar things have been done. Robot arms can be moved by the mind. Rats brains have flown F14 fighters. EEG sensors placed directly onto the brain (rather than onto the head) produce far more detailed information - it's not a stretch to suggest that some day a sensor layer will be placed onto the inside of the skull with a connection to the outside world.

You could, of course, play with EEG technology yourself. The OpenEEG project details the hardware needed and provides some basic software. See if you can find ways to get better results by changing the antenna used. See if you can get results good enough to be able to control your character in Quake or Netrek - a lot simpler than your average F14!

I mean, wiring your home computer to your brain... what could possibly go wrong?