Re:Scientists Closer to Reading Words From Your Brain
by treeface on Thursday September 09, 2010 @12:37PM
(#1675927)
That's some interesting research they're doing. I'd be be more interested to see if somebody can translate it to an electroencephalographic device. I suspect that placing electrodes directly onto the surface of the brain would be a dealbreaker for those of us without an immediate need to have our thoughts processed by a computer in real time (however cool that concept might be).
[ Parent
| on HN ]
Re:Scientists Closer to Reading Words From Your Brain
by sp332 on Thursday September 09, 2010 @12:43PM
(#1675955)
There's an external headset that "unwrinkles" your brain algorithmically and gets very accurate readings without requiring direct contact with your brain. http://www.ted.com/talks/lang/eng/tan_le_a_headset_that_read...
Re:Scientists Closer to Reading Words From Your Brain
by frisco on Thursday September 09, 2010 @01:29PM
(#1676116)
Hi, neuroscientist here. The stuff about it "unwrinkling" the brain is nonsense. Emotiv is a running joke in the scientific community, unfortunately. EEG has some fundamental problems as a control mechanism, and isn't interesting from
a brain-machine interfacing perspective. It's unfortunate because we all want to see neural interfaces, but it will either have to be invasive or come from a major Nobel-worthy innovation.
Re:Scientists Closer to Reading Words From Your Brain
by ramanujan on Thursday September 09, 2010 @10:55PM
(#1677726)
Hmmm, their demos look good. Are you saying the headset doesn't work at all, or not well enough, or not for a broad range of tasks?
Re:Scientists Closer to Reading Words From Your Brain
by frisco on Friday September 10, 2010 @03:40AM
(#1678193)
EEG measures mass synchrony between different areas of the brain. That's extremely hard to control consciously and takes a lot of practice. Even when you can do it, it's a 1-dimensional signal. It's not even clear you want to control it consciously; EEG therapy is used in psychiatry every now and then for treatment of a few disorders, and can lead to personality changes. The "inverse EEG" problem, which relates to the distortion caused by the skull, is basically computationally intractable (which is why ECoG is used). EEG is also totally drowned out by EMG (muscles) if you aren't careful, and it's nearly impossible to tell EMG and EOG (eye movements) apart from "real" EEG.
In Emotiv's specific case, though, these fundamentals are the least of their problems. They're three (four?) years behind schedule without anything really shipping even now, and the units that I have seen in the wild are of low build quality.