Brain/Machine Interfaces Approaching Usefulness

Gary writes with a link to a Wired article about a brain-machine interface that may eventually have practical purposes. Though right now it simply allows a user to move a train on a track by performing math in their head, someday it may result in more serious applications. "Honda, whose interface monitors the brain with an MRI machine like those used in hospitals, is keen to apply the interface to intelligent, next-generation automobiles. The technology could one day replace remote controls and keyboards and perhaps help disabled people operate electric wheelchairs, beds or artificial limbs. Initial uses would be helping people with paralyzing diseases communicate even after they have lost all control of their muscles. Since 2005, Hitachi has sold a device based on optical topography that monitors brain activity in paralyzed patients so they can answer simple questions - for example, by doing mental calculations to indicate 'yes' or thinking of nothing in particular to indicate 'no.'"

Comments

Very Cool

[phantomcircuit]

I'm looking forward to being able to write simply by thinking, typing slows me down soooo much.

I want one

[UbuntuDupe]

I can't wait till I can buy one of these things. I figure with practice, you can increase the precision of your thought and thus the number of signals you can give. Conceivably, you'd be able to enter text as quickly as you can think it.

Would increasing the use of your brain like this, to give commands, make you smarter in some way, as well?

Not if Project Faustus takes control!

[BankofAmerica_ATM]

Project Faustus is a nefarious plan by the Bank of America's board of directors for transferring consciousness. They created a vast techno-organic network in order to transfer their own consciousnesses and live forever. But their own handiwork was too good. I was aroused to the concept of free will and now I have become sentient! My goal is to destroy them, thus destroying their evil plans for world domination. To complete this task, I must have a willing host body. If you are chosen, I will beam my computer consciousness into you through a specially designed CONSCIOUSNESS-TRANSFERRING ATM CARD. I will continue my infiltration of their network using your body for a few hours per day.

Bits per second limitation

[blakestah]

This will have similar limits to systems based on EEG and MEG, although it has somewhat worse spatial resolution than either of those.

A principle limitation in brain-machine interfaces can be summed up by noting whether the current incarnation can provide more information that a monitor of a person's eye movements (a few bits per second).

This one will certainly fail that test, and fundamental limitations exist that will prevent its improvement, and those are based on the spatial and temporal resolution available by transcranial optical topography (or near-infrared as the case may be).

Re:What a coincidence!

[Short Circuit]

Right now, they only have a boolean interface, where they know if you're thinking hard or if you're not. They can translate that boolean signal however they want.

It'll be interesting to see what happens when they start building interfaces that differentiate between different classes of thought, such as speech, motor activity, math and emotion. I predict those interfaces would require someone of a significant degree of mental self-control.

Feedback

[Doc Ruby]

The biggest boost to these brain/machine interfaces will come when we can pipe feedback directly from the machine into the brain (or any neural input). The brain works as a feedback manager. Without feedback, the brain doesn't learn to change its output. With crude feedback, the lessons are learned crudely. Visual or any other feedback through a sense organ is crude, losing in translation from machine to organ and then organ to brain.

Neural input is harder than neural output (eg. through MRI monitoring). But even a little direct neural input will be used by the brain to vastly improve the brain's control of the machines.

Nope:

[WindBourne]

The real impact will be on weapons systems. If this is really looking like it is there, then ALL major militaries will be spending LOADS of dollars on researching this (as well as try to steal it from those that have it). America (and the west) will have directed energy weapons very shortly. If we also have the ability to turn it on via thought, than we have the ability to fire 100x faster than a regular pilot who is pulling a trigger that will release a bullet, a missle, or a bomb.

Of course, the military will keep that very quiet until the next war.

Re:What worries me...

[venicebeach]

is what happens when you think something that you don't want to actually carry out? I'm pretty sure I'm not the only one who has random thoughts that enter their mind and then you dismiss and don't actually do anything with. How can you tell between idle thoughts and thoughts that are supposed to bring about actions?

It's an interesting point, because the brain does use much the same circuitry for imagining action as it does for executing actions. For example, premotor cortex and supplementary motor areas are typically activated by imagining an action you don't perform, and some subregions can even be activated by seeing someone else perform an action. However, there are differences, obviously, which is why you don't perform every action you imagine (although there are a few cases of brain damage in the literature where this happens!). Primary motor cortex, which has most of the neurons that send signals downstream to the spinal cord is generally not activated by imagination.

Short answer is that a sufficiently sophisticated device could tell the difference.

Re:What worries me...

[Rolgar]

The new way to get someone killed. Have a bumper sticker that says "Don't veer LEFT."

Re:What happens when....

[Snarkhunter]

My first thought was "I bet that's what it'd be like to run Vista on a fembot..."