Brain-Control Gaming Headset Launching Dec. 21

An anonymous reader writes "Controlling computers with our minds may sound like science fiction, but one Australian company claims to be able to let you do just that. The Emotiv device has been garnering attention at trade shows and conferences for several years, and now the company says it is set to launch the Emotiv EPOC headset on December 21. PC Authority spoke to co-founder Nam Do about the Emotiv technology and its potential as a mainstream gaming interface." One wonders what kind of adoption they expect with a $299 price tag.

Re:Bypassing normal I/O mechanisms of the brain

[MrNaz]

"Given the little knowledge I have of neural network functioning"

It seems it's even littler than you think.

Re:NIA out for a year.

[Anonymous Coward]

This one likely won't work either - or, at least, it'll just do EMG (electromyography) and pick up jaw clenches and blinks and whatnot.

EEGs are so contaminated with noise (both from the brain and from muscle movement) that pulling out single events is tricky. Studies that analyze EEG data usually have to average a lot of events together to get rid of the noise (ERP). There are a few ERP components (scalp potential changes) that are pretty well-known and usually prominent enough to maybe get on a single-trial basis (e.g., P300), but those easy-to-detect ones are pretty much all-or-none - in other words, no finely graduated control like an analog joystick.

I sorta read the article and the guy claims that this will 'find the sources in the brain' where activity originates. Source localization with, what, 8 electrodes, half of which are occipital? Mmmmyeah, that'll work great.

Save your $300 and buy a nice joystick.

Re:Could be a half-decent toy, if priced well

[uglyduckling]

You can definitely record useful EEG data on neonates using forehead electrodes without gel. It's called a Cerebral Function Monitor [natus.com]. It looks like there is research in using it with adults [ccforum.com] too. These are mostly used in patients that are comatose or (medically) paralysed, so I suspect there would be issues with interference from motor nerve signals although these would have a very different pattern so I suspect could be filtered.

Misleading

[Jeffrey_Walsh VA]

Should be brain-controlled

Other Applications

[smitty777]

I'm a bit surprised (or maybe not) that the focus of the discussions has been on the gaming aspect of the device. I know it's not perfect, there are a lot of bugs to work out, and it's been around for a while, but I can see tremendous application and potential for this technology. This could provide quadriplegics with access to software, allow another interaction pathway for those with their hands occupied on critical tasks (pilots, surgeons, police).

I wonder how much the hand gestures were required to move the objects? I'm sure it's a way to "cognitively prime the pump" at this point, but could it be done without the gestures? Or could someone learn to do it without them?

Re:Could be a half-decent toy, if priced well

[Beefmancer]

As someone *else* familiar with BCI research, I can back that up, $299 would be a good deal if the software is any better than openEEG. Buyers should be aware however that their website makes some really far-out claims for the device, and that the marketing is realllly shady. Plus, as I've pointed out here before, I've heard firsthand from one of the biggest names in the field that he met with the President of this company who failed to demonstrate that this device could do anything. Even the best in academia can get only wobbly control in 2d... thus the odds are very low that their 2d cursor control, even with tons of practice and a bald head, is going to be good enough for shooters.

To give them credit though, it might be good enough for pac-man! And if done properly it could get kids excited about neuroscience.