Brain Controlled Computing a Reality

pchernyakov writes "Cyberkinetics Neurotechnology Systems told attendees at the American Academy of Physical Medicine and Rehabilitation annual conference that a 25-year-old quadriplegic with wires run from 100 tiny sensors implanted in his brain and out to a computer can use his thoughts to control a computer well enough to operate a TV, open e-mail and play Pong with 70% accuracy."

Re:How about....

[HBI]

I keep thinking that a 'human bus connector' would be a good idea, once you've identified the areas of the brain most conducive to electrode implantation for control purposes, you create a standard connector and tolerances for the controls. Identify motion axes that can be trained. Create a computer that hooks into the bus connector (mounted in the most logical place, perhaps behind the neck?) and allows the user to train using the motion axes in a therapy environment, then move them up to the vehicle that can provide mobility, a grasping hand and communications.

The advantages of this would be that as new hardware is invented, the brain electrodes wouldn't need to be re-implanted and the new hardware could simply take advantage of the existing control interface.

It's been a dream to regrow spinal cords. This provides a technological end run that while not 100% desirable, gives them a far more mobile and productive existence than would otherwise be possible.

Re:yikes!

[StuckInSyrup]

But on the other side...some quadruplegic men are able pleasure women. It requires a number of intact nerve strings and the hormones do the rest. Howewer, they are not able to "feel" the act. And the woman has to do all the "work".

Re:Original thought keyboard?

[Eschatus]

Are you thinking perhaps, of http://www.ibva.com/ [ibva.com]?

Re:RTFA PLEASE...

[StateOfTheUnion]

Hardly true . . .scanning 100 neurons does not generate 100 degrees of freedom. That would imply that each neuron is a discrete controller that can be stimulated at will by the user. Neurons works in conjunction forming netoworks through axons and synapses and fire in combination with one another. They are not independent control elements. The user is probably firing off a pattern of neurons at a time and this pattern is picked up by the electrodes and interpreted by software. Repeatable neural patterns that are able to be produced on demand are then linked to inputs on the TV, in the pong game etc.

In fact if you had actually paid attention and thought about the article after reading it, it would be rather obvious that the quote

There are 100,000 neurons in a square millimeter of cortex. There are very precise codes in the neurons. The details matter."

Is referring to the details of neural patterns being picked up, not individual neurons. Just as the quote from the article implies, the devil is in the details.

Re:How about....

[Anonymous Coward]

I know it's a tremendous drain on your all-important time, but had you actually READ the ARTICLE, you would know that this amount of control isn't currently feasible.

Even in the bloody summary it said he can only play Pong with 70% accuracy - that should be enough information to allow you to deduce that the control isn't exactly tip-top.

Perhaps I'm expecting too much.

Re:What does Captain Pike think?

[teromajusa]

Did you read the article? Each sensor monitors the activiation of 1 neuron. But the state of 100 neurons (the limit of the current device) is not sufficient to guage what the brain is trying to do. We don't need smaller sensors, we need alot more of them.

Re:The Matrix

[lee7guy]

Start practicing [bbc.co.uk] today.

This has already been done - 5 yrs ago

[sdfan]

This is nothing really new. Dr. Phil Kennedy was the first to successfully implant a person at Emory University in 1999. He has done five more implants and two lasted over two years. Go here for more details [emory.edu] His company is working on the third generation version. Neural Signals [neuralsignals.com]

Re:Birth of Cybornetics...

[HalfFlat]

[...] and more significantly, neurons have a steady rate of death, and zero regeneration.

That's not strictly speaking true - while it was long held that the neurons you have at twenty are all you are going to get, it's been discovered that we do in fact continue to grow more neurons over time. At least, in some parts of the brain. Whether or not neurogenesis occurs in the neocortex of adult primate brains is still a matter of dispute.

Still in the end, it appears to be a losing battle.

In the meantime, if you want to encourage neural growth, keep stretching that brain. Learn new stuff, do new things. Don't stop. Drink alcohol, but in strict moderation. Oh, and don't smoke, and stay aerobically fit.

Re:What does Captain Pike think?

[figurewmeat]

Working out the kinks is a critical step indeed. I interned at a bioengineering lab this summer, and a huge problem was just keeping channels active. The brain sloshes around inside the skull quite a bit, and to have a stiff electrode intrude on this can be problematic to say the least (if a jet pilot with an implant were to pull a few Gs I imagine some sort of self lobotomy).

Next are immune reactions to foreign objects in the body. Coatings over the electrodes can help, but are not a guarantee.

Finally, these electrode arrays are pretty large. Technology can improve this, but imagine invasive brain surgery every time you need to upgrade, or being stuck with the 1st generation mind-link ipod for life.

I'm waiting for nanomachines to solve many of these problems. When (and if) I make it to grad school perhaps they'll have it set for me... or maybe I'll volunteer as I'm never going to fly a jet.