Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Medical Brain Computer Brain interface, anyone?

  1. Feb 10, 2007 #1
    There seems to be lots of interesting news about Brain Computer Interface recently (and not so recently).. for example artificial vision (even if only a handful of pixels) and controling prosthetics.. also some interesting articles on rat-neuron computer chips etc.

    What I find most interesting is that it seems to be the organic side that is doing most of the work. I mean we are connecting neurons to computers, but the organic side is the one figuring out how to interpret the interface.

    Has any experimentation been done with connecting brains directly? Is it possible that by creating a comparatively small number of crosslinks between two rat brains in the appropriate lobes they could start sharing knowledge?

    With sufficient numbers of connections, would they experience just one conciousness (the way our two hemispheres seem to, for example)?
  2. jcsd
  3. Feb 11, 2007 #2


    User Avatar

    mind meld

    Let me get this straight; you want to share consciousness with a rat? Oh, two rats, I see. Well one thing that works in your favor in the structure of the brain, ie that it brings all the signals out to the surface to do the processing (well, for the higher order and sensory stuff at least) and it tends to maintain the proper mapping instead of scrambling the bits of sensory input all together. So if you made a membrane that laid over the surface of the brain and connected to another membrane with sufficient channels you could, in effect, connect the two brains together. However, in practice there are a number of problems that arise. First, there is a high probability that the two rats will think they have gone schizophenic and not realize what is happening. The stress and panic spreads to the other and gets fedback so they both go into a panic attack. Or you try to hook a male and a female and the male brain gets too confused by the indecisveness it's suddenly experiencing and it goes berserk. Two females works fairly well, each has such ego strength to not be too disturbed by the other but the flightyness is doubled so there is nothing worth exchanging. That's what happened here when we tried it anyway.
  4. Feb 11, 2007 #3
    er.. very good.. You do realise that 'thought experiment' doesnt mean something you just made up.. right? :grumpy:
  5. Feb 12, 2007 #4


    User Avatar

    Pray tell, what do you think "thought experiment" means? If you find fault with my results, get down to specifics. You're not a scientist are you?:rofl:
  6. Feb 12, 2007 #5
    er.. actually I am a scientist, but my topic was blunting pointy objects, not ratology. right.. have you um, published?
  7. Feb 12, 2007 #6
    RDX? Hey, man - good to see you on PhysicForums...I have to say that I'm a little confused myself..."That's what happened here when we tried it anyway."...are you implying that you've <i>actually</i> done this? Or have I misread you?


    This seems like a very interesting concept...I actually plan on working on computer/brain interfaces in the future. I must admit that I never really pondered direct brain to brain connections, because I don't think it would work on a general scale, as "knowledge" seems to be a semi-global pattern of brainwaves that only makes sense in the context of personal history. I don't think there is any way for a one brain to really probe the memories of another - it would simply be perceived as noise, and the two brains would probably interfere with each other in a chaotic manner.

    <i>However</i>, I can't help but wonder what would happen if you had a one way connection. Consider this: Two neuromorphic chips, one on the motor cortex of each rodent, with proper connections between the two chips. The signals from the axons of the motor cortex of the "intentional rat" (R1) are amplified and relayed to the motor cortex of the "controlled rat" (R2). My guess is that R2 would begin moving erratically, and his brain would most probably regain control. (You might have a sort of tug-o-war.) But what if you cut the axonal connections of the interneurons coming into the cortex of R2? Basically, you would have a limp rat (R2), that is essentially the <i>same</i> as a robotic apendage that we know monkeys are able to control. Over some time, it's possible that R1 would learn to control the body of R2, albeit very crudely. I don't know if rats are the best method for this kind of testing, and I should point out that I'm only just beginning my education, so my questions are rhetorical and probably fundamentally flawed.

    But I'm intrigued - Someone with some knowledge, contribute!!
  8. Feb 12, 2007 #7
    Oh, html is filtered out, huh? <b><u><i><font face="engravers mt">damn</font></b></u></i> (Kidding)
  9. Feb 12, 2007 #8
    basically you use the [] symbols instead of <> eg duh! o:)

    (edit: MISTAKE! I thought the bionic arm was training new muscles.. the article implies they find the original nerves, so some of my argument is not supported by that link. There are many other examples of how adaptable the brain is however)

    Some interesting links:

    Woman with bionic arm. It connects to chest muscles rather than directly to the brain, but it is still bandwidth, so who cares. What is interesting to me is how the woman can also feel through this mechanism. with practice it all becomes natural because the brain simply learns to reinterpret.

    "Brain in a dish" a rat-neuron computer used to fly an aircraft simulation. So we can do reasonable complicated things without understanding the entire brain.

    The wikipedia for a bunch more links

    My belief, based on evidence such as above, is that the brain is very good at redesigning itself to make use of informaiton given to it. (even right now, typing, I barely think in terms of moving my fingers. the sensation is more like pouring text into this edit box.) and rat neurons flying a plane? that wasnt part of the original ratty design!

    So what if the womans muscles had been attached to a keyboard instead? (for example 8 switches).. this is enough for a byte of information, eg a character of text. I bet she could learn to send and recieve text totally naturally, with no sensation she was actually twitching chest muscles. She would have an entirely new 'TXT' sense. Immediately you could have a form of telepathy-through-email, or control devices such as organisers or calculators.

    now how about these scary links:

    How about that first one which mentions removing one hemisphere as a solution to epilepsy? Or the second link with the famous description of how the hemispheres can be severed with the result that in some sense they become independent.

    The point is that the brain is very robust because of its parallel nature, and ALREADY is comfortable to coordinate with another intelligence so seamlessly as to appear as a single conciousness.
    Last edited: Feb 12, 2007
  10. Feb 12, 2007 #9


    User Avatar

    I agree with you conclusion that coupling brains would result in chaos. And that it is a pointless endeavor. That was what I was saying in my own confused and irrational way. I leave rats to themselves and experiment on cats, which have no perceivable brain function beyond sensory processing. I imagine hooking up two cats that way and when one decided to move, the other would start to too, resulting in spastic chaos. I see it like superimposed images like in the old days when you could forget to advance the film in a camera and take two pictures.

    I have a basic philosophy that we should not compete against machines in the areas where machines are better, faster, stronger. And I work to make machines process information better. Regarding that article you sent to me, they talk about trying to enhance human senses, but positing that the other centers have adapted to what is extant so enhancements are lost. You have to rebuilt every system, by which time you don't really need the human anymore. I see synthetic intelligence surpassing humans in decades in every realm like it has in chess already. There need to be advances in the structures of hardware, current computers are ill suited to the "computations" required, but I don't see FETs as neural analogs.

    Anyway, sorry for the confusion; no rats were injured in the making of this tirade.
  11. Feb 12, 2007 #10


    User Avatar

    ...second wind...

    Now let me comment on the second part of what you said, about a oneway connection between animals. I see such remote control to be theoretically possible IF there is an excellent transmission of sensory information back to the controller rat. If you have read my article on synthetic intelligence you may note that sensori-motor is unified in a tightly coupled pair, based on reflexed at the lowest level. Now reflexes don't go through the brain and would not be available to the controller rat. However if enough proprioceptor data is available, the controller could adapt in time. Whether this new training would occur to the detriment of the initial motor training or in addition to it is an interesting question. But I don't really see the point in such abilities.
  12. Feb 12, 2007 #11
    Hey at least youll not get lost in the modern maze that is the supermarket, you may get stuck in the dairy section, but you know that somehow youll find your way out. :rofl:
  13. Feb 12, 2007 #12
    Hi.. I wasnt expecting automatic knowledge transfer or compatibility, or syncronised brainwaves, but that perhaps the brain could learn to interpret this input so well that it does not notice the act of interpretation. This seems almost foregone to me since this already happens with all our senses then the point is could this communication happen faster than we can compose speach?

    Could you send me some links or something to put me into context on your opinions?
  14. Feb 14, 2007 #13
    Sorry, B.E.M., I've been occupied. I wish I could furnish you with links, but I'm afraid my knowledge comes from the biological aspect of BCI's, not from BCI's themselves. I don't know how much you know about the brain, so I guess I would recommend Joseph LeDoux, (I got my name from one of his titles, actually), Walter Freeman, Elkhonon Goldbergh...to name a few authors (aimed at the masses, I'm afraid.)

    Freeman describes something called Solopsistic Isolation...basically it describes (I believe) why meaningful knowledge couldn't be shared. As for the possibility of "sharing consciousness? You would need sort of a biological hub to connect all four hemispheres, or an electronic equivalent of the corpus callosum - and that's just too speculative for me to comment on. Besides, I don't see any good coming from this - if you have any ideas besides information sharing, which I can't see happening, please share.

    As for links...I did find some videos corresponding to some experiments that I've read about in the past, on YouTube. This video shows the infamous monkey/robotic arm experiment - although "Reenactment" appears on the screen, and this concerns me.

    This one is a far more impressive video, using a non-invasive BCI to control a humanoid robot. However, do note that BCI's in the past have been very crude...they generally are limited to simple directional commands from the brain, and they must be trained accordingly. The complexity of this demonstration worries me; I can't help but wonder if the "walking" functions are hardwired into the robot (or interface) itself. I find it hard to imagine each step resulting from a direct brain command.

    Finally, brainloop video: This is the most promising application - direct software interaction. I see huge potential for this type of technology, and not too far in the future (as opposed to remote/brain controlled robots...although until AI does surpass us, the robot idea could come in handy for probes and unmanned vehicles.) If you watch the screen that the subject is watching, you can get an idea as to how these BCI's work. You can see that he is basically thinking "move hand/arm right/left, etc." This may only be a "limitation" for more mature minds. If a young child was brought up in an age of BCI's, his mastery would most probably be far greater.

    Thank you for the links, B.E.M. Although I've seen some of this, I hadn't yet heard of the rat brain in a dish. I'm actually going to be starting a thread about this very soon.
  15. Feb 15, 2007 #14
    I did find this link,

    I didn't really understand it, but it did make me realise I am totally out of my depth there. However this is the reason why I phrased my original post as a question, ie is there any work happening in brain-to-brain interface.

    lets forget I mentioned CONCIOUSNESS!

    The only reason I expected there might be simple progress is a series of anecdotes we have all heard about how adaptible the brain is to learn how make sense of new inputs. eg
    • A blind person whose sight is restored has to be trained to see
    • If you wear goggles that put what you see upside down, after a while
      your brain realises how to compensate and you cease to have trouble (until you remove the goggles

    I think humans especially always learn ways to communicate, for example, imagine two twins who from birth had the left hand of one and the right of another trapped in a box, able to touch only the other hand. I would expect them to develop some sort of language of touch to exploit this. BCI just seems a less clumsy way to consider such an experiment.

    The reason I think it is so interesting is because it is near future technology that could allow groups to act like entirely new species, with entirely new forms of group organisation.

    How would two people behave if they automatically shared the same emotions and pleasure/pain?

    What if a group elected to share a private form of communication through which it was impossible to lie?

    If BCI develops to the level that it is faster for entering/recieving data than keyboards and speach, and more convenient, then what will communication between people feel like?
  16. Feb 16, 2007 #15
    I see where you're going with this, B.E.M., and I agree. Ray Kurtzweil wrote a bit about this, but I wouln't point you toward him if you're looking for related works. He's a futurist - his job is to latch onto the work of others and speculate...and incite undue optimism in people by exploiting our natural fear of death and isolation.

    Brain to brain communication would be consciousness sharing - don't trivialize it, just recognize the intricacies. For now, I'll just try to keep informed...I'll pass whatever I find to you, if you promise to recipricate.

    Perhaps instead of brain to brain communication, it might be easier to set up a standardizing interface. For instance, each member of a "group" would have some sort of synthetic neural network that has specifically adapted to their brain, and then seperately those interfaces could be combined as a LAN, so to speak. In this way, information could be shared that is much more meaningingful than spoken language, because one's vocabulary wouldn't limit one's expression. It could be translated and mediated using a standard (as computers communicate currently). This would improve communication because there would be less room for interpretation. What one means to convey would be conveyed, because translation in both directions would be executed by a device that has a unique, symbiotic relationship with the individual party.

    Although we may be able, in the future, to directly adapt to others to form a group without such standard, it would take a long period of training. This training would have to be performed with each and every change the structure of the group, or if you wanted to switch groups entirely. Not to mention over time, this close relationship would drift -according to both Freeman's model of thought and the schema model of memory. One relationship, one standard.

    RDX: if you're still with us. I did read your article on synthetic intelligence (provided you're referring to "Synthetic Intelligence Revisited". You must remember that 1) we are on two very different planes of understanding (this is a compliment), and 2) I am not a computer scientist. Frames and stacks? I can't tell if I agree or disagree...do you have a less technical description of your model that I might be able to read?
  17. Feb 16, 2007 #16
    Last edited: Feb 16, 2007
  18. Nov 22, 2008 #17


    User Avatar

    I have been thinking something along almost the exact same lines as B.E.M. for many months now, and recently found this page with a google search.

    I think at the very least, this should be tried. One cannot just dismiss this idea after 10 mins doing "thought experimentation." Biology is constantly surprising us, especially the brain. We really have no idea if brains can coordinate directly or not because we just haven't tried it yet.

    I would hypothesize (just hypothesize) that if you completely covered the brain in sensors and wired two mice together they would eventually become a single thinking unit. Because: the brain is extremely adaptable, the two halves of our brains appear to be able to do so already, and functional layout of the brain is at least somewhat consistent from organism to organism. For instance, scientists have put a BCI on a cat's visual cortex and found that position in the cortex roughly correlates to the position an object is seen at (like, activity at the top of that brain area corresponds to what's being seen at the top of the cat's field of vision). I do not think the rats could immediately share "meaning" but would have to learn how to, possibly from a very young age. Another important thing might be to gradually connect the brains, i.e. not open all the channels at first, or attenuate all the signals, so that (as you guys have been saying) the animals would not break down in utter confusion.

    Thank you everyone for the interesting articles so far. I'll be sure to post anything I find, too. I'm actually considering trying to look into this as a career :)
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?

Similar Discussions: Brain Computer Brain interface, anyone?
  1. The brain (Replies: 1)

  2. The brain (Replies: 2)

  3. The Brain (Replies: 23)