Re: brainwaves - NOT
Wed, 5 Aug 1998 15:20:30 EDT

In a message dated 8/5/98 7:53:13 AM, wrote:

>While looking for something entirely different, i stumbled aX this -
>Believe it or not - and I'd _really_ love to hear a testimonial from
>someone, it's an input device (and apparently commercially available, to
>boot) which is triggered by brainwaves. Comes with a set of Hypercard
>XCMDS, which (unless I'm much mistook) should also be usable in

I have some acquaintance with EEG signal processing, brainwave coherence
analysis, and attempts to correlate these to anything like cognitive states or
mental command. I believe it's possible in theory, and maybe in commercial
practice, but I'm unaware of ANY (non-classified anyway) success stories.
Obviously, if/when this IS achieved, a true mind-brain-machine interface would
have tremendous and far reaching implications for technology, communications,
media and civilization itself!

I've experience this IBVA system demo'd several occasions over the past
several years.
I have never been able to see ANY response modality that was:

(a) reliably repeatable; and
(b) repeatably correlated with ANY form of conscious intent / user cognitive

The most I've seen is an experienced biofeedback practitioner acheiving a
"general direction" of response flavor over several minutes drift; nothing
like discreet event evoked responses or feedback.

Thus, the overall brain wave (EEG) state of affairs, with some post-
processing, simply is fed as an input into various media results. I have
never seen it be more than, from a human factors standpoint a generally
chaotic or almost random influence. Overall EEG frequency-related data WILL
affect output, but merely that, is NOT A USEFUL nor reliable
input/control/feedback methodology.

Maybe IBVA's technology has improved since my experiences... anyone out there
on the list have another opinion? If it had, I'd think we see it in the press,
extensive scientific literature citations, etc. etc. and we haven't as yet.

My guess we'll have to wait for room-temperature superconductors or fiber-
optic interferometric-based brain magnetometers, as cortical-surface EEG data
appears to crude and noisy. Even with some future superconductor-based /
high-powered brain 3D magnetic field sensing and high-powererd signal
processing, to correlate to cognitive states, perception, memory, cognitive
state, or user intent is A VERY TALL ORDER INDEED. Might even require future
holographic neural network architectures, etc. and extensive supervised
training program on large populations of humans to achieve ANY degree of
effective inter-class classification (of cognitive states) in face of extreme
intra-class ambiguities and artifacts.

As to Christmas list? try a joystick, or... ?

David Clark
Chief Technology Officer
FreeSpace-Interactive˙ Technology, LLC