Wireless connection to brains?

Stephan Anagnostaras stephan at nospam.ucla.edu
Tue Dec 1 06:16:05 EST 1998

In article <3663C466.14CC at ntc.nokia.com>, Janne.Mantykoski at ntc.nokia.com wrote:

> Hi,
> I am interested to get information concerning brain research. I am aware
> that brain activities can be measured already with current devices.
> Doesn't it mean that it has been also possible to make a connection
> between the formation of words in our mind and the results of
> measurement of brains?
> Janne
> Janne.Mantykoski at ntc.nokia.com

Not really, the signal is too noisy... the way you can do this is to just
do it empirically. This means you don't actually know what you're doing,
but you take a bunch of measurements (like EEG -> ERPs) and then correlate
them with some known values (i.e., words the person is thinking or
whatever). Then you just create a lookup table for some particular set of
parameters which correspond to the "thought."  

Unfortunately, most things are only weakly correlated with some particular
set of parameters and you can't get something like a specific word, more
often you get something like "surprise" (which is the most reliably
measured thing). Plus you can imagine the computer power requried to do it
if the lookup table gets large and the number of parameters requried to
discriminate well is large.

Now all this said, it doesn't mean you couldn't find a particular set of
measurements that are reliable and exploit them as a command system. I
know for sure Microsoft is spending money on something like this.


More information about the Neur-sci mailing list