Maybe we should allow extra cyberlimbs..., Swedish study shows the brain might be able to handle a third arm Options

[Tyro]

Apparently our brains aren't limited by the number of limbs we're born with:
http://www.thelocal.se/32234/20110224/

[ggodo]

THIS IS AWESOME! I WANT A PREHENSILE TAIL!

Did I say that aloud?

[SpellBinder]

And here I thought it was going to be more than just an illusion to the human brain. Interesting none the less, but I was honestly expecting a bit more.

[phlapjack77]

My Swedish coworker is being suspiciously quiet about this...I'm a little upset, he must be trying to keep all the third arms for himself.

<eyes coworker suspiciously>

[Faraday]

My Swedish coworker is being suspiciously quiet about this...I'm a little upset, he must be trying to keep all the third arms for himself.

<eyes coworker suspiciously>

Maybe he got an extra third leg and is embarrassed. :x

[phlapjack77]

Maybe he got an extra third leg and is embarrassed. :x

I'd rather not find out (IMG:style_emoticons/default/smile.gif)

not that there's anything wrong with that...

[MJBurrage]

This was already known from prosthetic research on primates.

In the experiments, the primates had sensors implanted into their brains to record the signals that controlled their natural arms as they used a joystick to control a robot arm.

Step one, the sensors are used so that a computer can learn what their motor control signals look like in the brain.

Step two, the joystick is disconnected, and the robot arm's commands are actually sent by the computer based on what the sensor sees in the primates brain.

What was unforeseen, was that the primates figured out that the joystick was no longer connected, and that they could move the robot arm just by thinking., so they stopped bothering to manipulate the joystick.

The end result was that they could then choose to move their natural arms separately from moving the robot arm. IE although the robot arm was not physically attached to their body, their brain was still controlling three arms at once.

[Sengir]

My Swedish coworker is being suspiciously quiet about this...I'm a little upset, he must be trying to keep all the third arms for himself.

<eyes coworker suspiciously>

Hmmm, didn't a major part of the Chernobyl cloud drift over Sweden?

[Tyro]

This was already known from prosthetic research on primates.

In the experiments, the primates had sensors implanted into their brains to record the signals that controlled their natural arms as they used a joystick to control a robot arm.

Step one, the sensors are used so that a computer can learn what their motor control signals look like in the brain.

Step two, the joystick is disconnected, and the robot arm's commands are actually sent by the computer based on what the sensor sees in the primates brain.

What was unforeseen, was that the primates figured out that the joystick was no longer connected, and that they could move the robot arm just by thinking., so they stopped bothering to manipulate the joystick.

The end result was that they could then choose to move their natural arms separately from moving the robot arm. IE although the robot arm was not physically attached to their body, their brain was still controlling three arms at once.

Nice! Source? Not that I disbelieve; I just want to read the article and see pretty pictures ^_^

[eyeBliss]

The study this post is based on is just a disruption of one of the internal frames of reference the brain uses to organize things. Making the brain believe that an arm is part of the body is a far cry from being able to manipulate an additional limb. You can create the out of body/near death perception with targeted electrical stimulation to certain deep brain structures (a French neurosurgeon who's name escapes me documented this), this doesn't mean that the brain has the capacity to control an additional phantom body (but you could probably synch your motor output to multiple selves if your have some sort of identical kick line of yourself fetish). Here are the challenges as I see them:

1. Need to remap both primary motor and sensory cortex to accomodate the new limb
-you will be overwriting somatotopic maps that are optimized for other structures, losing functionality of those structure and not ideally accomodating the needs of the new limb.
2. Cerebellar error correction for the new limb will prove problematic.
3. How do you code vectors in pre-motor cortex for a limb that didn't exist until right now?
4. You need to build in the necessary spinal reflexes for the new limb.
5. You're going to need a good attachment point for the new limb and the associated muscle and bone structures to orient and control the limb (not my specialty)

The University of California, San Franciso (and some allied institutions) have done some interesting things with using direct output from human motor cortex to direct a robotic arm if anyone wants to look into it further.

Other fun facts: Part of the trick with the third arm (synching two congurent somatotopic sensations) can be done with two non-analogous body parts and is first mentioned in indian erotic texts that are about 2000 years old.

[Makki]

I wouldn't know where to put it...

[Kyoto Kid]

...man having a third arm would be so bloody handy. I think of all the times I arrive at home carrying two bags (made of paper as they banned plastic grocery bags where I live) of groceries and need to unlock and open the door while the rain is bucketing down. Would be nice not having to worry about putting a bag down on the wet landing only to have the bottom give way when I pick it up.

Other advantages,

I could actually carry a bumbie, a cup of coffee, and have my transit pass at the ready to show the driver without having to put something down and fumble through my wallet thus holding a busload of people up.

I could signal a turn while riding a bike without taking a hand off the handlebars,

I could tend to several pots on the stove at the same time.

I'd get my CG workstation built.

I could run the shipping station at the day job 50% faster.

I could fly a dual/quadline sport kite and drink a beer at the same time.

I could play the Tocatta Adagio and Fuge in C Major (for organ) by JS Bach on the piano (love that piece).

I'd be wicked on a drum kit.

I could use both a Vuvuzela and "bangers" at the same time during a sporting event and be really annoying (well maybe this one's a disadvantage). (IMG:style_emoticons/default/wink.gif)

[phlapjack77]

...man having a third arm would be so bloody handy. I think of all the times I arrive at home carrying two bags

With a third arm, you'd end up trying to carry 3 bags home instead (IMG:style_emoticons/default/smile.gif)

I could use both a Vuvuzela and "bangers" at the same time during a sporting event and be really annoying (well maybe this one's a disadvantage). (IMG:style_emoticons/default/wink.gif)

It's really strange, I actually like the vuvuzela, both on the tv and in person. I think it adds alot of atmosphere...

I could fly a dual/quadline sport kite and drink a beer at the same time.

This is the single best reason to get a third arm - you can continue doing whatever it is you're doing, and drink a beer at the same time.

On the gripping hand, the Swedes probably won't share this tech with anyone else

[Tyro]

...man having a third arm would be so bloody handy. I think of all the times I arrive at home carrying two bags (made of paper as they banned plastic grocery bags where I live) of groceries and need to unlock and open the door while the rain is bucketing down. Would be nice not having to worry about putting a bag down on the wet landing only to have the bottom give way when I pick it up.

So get reusables. They're cheap and last a good while. I don't know where you live, but around here Trader Joe's has good ones.

[MJBurrage]

These monkeys have three arms

• Duke Medicine News and Communications (May 10, 2005) "Monkeys Adapt Robot Arm as Their Own" DukeHealth.org
• Carey, Benedict (May 29, 2008) "Monkeys Think, Moving Artificial Arm as Own" New York Times
• Ransford, Matt (June 3, 2008) "Monkeys Work Robotic Arm" Popular Science
• Brain–computer interface at Wikipedia

P.S. The wikipedia article also has an image from an experiment where a brain implant in a cat recorded what its eyes were seeing. Once we know the encoding the brain uses for vision, we can start working on cybereyes.

[Kyoto Kid]

So get reusables. They're cheap and last a good while. I don't know where you live, but around here Trader Joe's has good ones.

...not always an option as I often stop off to shop on the way home from work.

...and as I'm on transit and don't drive, can't just leave a couple in the back seat.

[Vegetaman]

This reminds me of the zombies that show up in Episode II of Wolfenstein 3D for the PC... The guys that have two arms with meat cleavers and a hand in the middle of their chest that holds and shoots a pistol. Plus they're like stealth masters.

[Tyro]

...not always an option as I often stop off to shop on the way home from work.

...and as I'm on transit and don't drive, can't just leave a couple in the back seat.

There are bags with drawstrings that fold up very small but can be unpacked to full size. My mom carries one in her purse.

[eyeBliss]

These monkeys have three arms

• Duke Medicine News and Communications (May 10, 2005) "Monkeys Adapt Robot Arm as Their Own" DukeHealth.org
• Carey, Benedict (May 29, 2008) "Monkeys Think, Moving Artificial Arm as Own" New York Times
• Ransford, Matt (June 3, 2008) "Monkeys Work Robotic Arm" Popular Science
• Brain–computer interface at Wikipedia

P.S. The wikipedia article also has an image from an experiment where a brain implant in a cat recorded what its eyes were seeing. Once we know the encoding the brain uses for vision, we can start working on cybereyes.

In all of the experiments listed, the arm of the monkey is restrained when they are manipulating the robot arm, so they only have two functioning arms. Motor cortex codes for vectors rather than individual muscles movements, so it works out nicely for a set up like this. You record while the monkey is orienting the arm, reconstruct what the monkey wants to do (I've done a bit of neural reconstruction work with Dr. Chafee at the University of Minnesota, but looking at some very different questions), and then program your peripheral to understand the vector coded. My point was that adding additional capacity (the third arm) presents a unique problem because you can't use existing neural infrastructure to do it. Regarding the cyber eyes, implantable micro-photo diodes already exist (I think most of the work was orginally done in patients with macular degeneration, but I don't have any citations on had...I could probably find them if anyone wanted...or any wierd neuroscience stuff if anyone is interested.)

[Tyro]

In all of the experiments listed, the arm of the monkey is restrained when they are manipulating the robot arm, so they only have two functioning arms. Motor cortex codes for vectors rather than individual muscles movements, so it works out nicely for a set up like this. You record while the monkey is orienting the arm, reconstruct what the monkey wants to do (I've done a bit of neural reconstruction work with Dr. Chafee at the University of Minnesota, but looking at some very different questions), and then program your peripheral to understand the vector coded. My point was that adding additional capacity (the third arm) presents a unique problem because you can't use existing neural infrastructure to do it. Regarding the cyber eyes, implantable micro-photo diodes already exist (I think most of the work was orginally done in patients with macular degeneration, but I don't have any citations on had...I could probably find them if anyone wanted...or any wierd neuroscience stuff if anyone is interested.)

First, wow. I love Dumpshock - we get the most interesting people here, with all sorts of useful knowledge.

Second, way to burst my bubble (IMG:style_emoticons/default/nyahnyah.gif)

[MJBurrage]

In all of the experiments listed, the arm of the monkey is restrained when they are manipulating the robot arm, so they only have two functioning arms. Motor cortex codes for vectors rather than individual muscles movements, so it works out nicely for a set up like this. You record while the monkey is orienting the arm, reconstruct what the monkey wants to do (I've done a bit of neural reconstruction work with Dr. Chafee at the University of Minnesota, but looking at some very different questions), and then program your peripheral to understand the vector coded. My point was that adding additional capacity (the third arm) presents a unique problem because you can't use existing neural infrastructure to do it. Regarding the cyber eyes, implantable micro-photo diodes already exist (I think most of the work was orginally done in patients with macular degeneration, but I don't have any citations on had...I could probably find them if anyone wanted...or any wierd neuroscience stuff if anyone is interested.)

It has been a while since I read the actual papers (rather than the summaries linked above), but as I recall, the monkies natural arm was restrained during the training process, but that once that was done, restraining was no longer required. I.E. the monkeys could control the robot arm without moving their natural one. Unfortunately, my copies of the scientific papers are in storage, but I my be able to get at them in a week or so.

As for vision implants in humans, I am aware of three methods that have been tried.

• A chip that stimulates the retina (for patients who have occlusions that prevent light getting to it naturally). As I recall the subject could make out general shapes afterward, but not enough detail to identify specific subjects.
• A cuff that wraps around the ocular nerve, and stimulates it directly. More difficult, since some visual preprocessing is done in the eye/retina itself. I do not recall the results of these experiments.
• A direct signal sent to the visual cortex. This has only been done on two or three patients in Europe, and had incredibly limited "success" (one of the subjects could detect relative light levels, but not general shapes, at the cost of severe migraines) As I recall, this research was considered unethical in the US due to the invasive nature of the brain implant related to the expected outcome.

[eyeBliss]

It has been a while since I read the actual papers (rather than the summaries linked above), but as I recall, the monkies natural arm was restrained during the training process, but that once that was done, restraining was no longer required. I.E. the monkeys could control the robot arm without moving their natural one. Unfortunately, my copies of the scientific papers are in storage, but I my be able to get at them in a week or so.

As for vision implants in humans, I am aware of three methods that have been tried.

• A chip that stimulates the retina (for patients who have occlusions that prevent light getting to it naturally). As I recall the subject could make out general shapes afterward, but not enough detail to identify specific subjects.
• A cuff that wraps around the ocular nerve, and stimulates it directly. More difficult, since some visual preprocessing is done in the eye/retina itself. I do not recall the results of these experiments.
• A direct signal sent to the visual cortex. This has only been done on two or three patients in Europe, and had incredibly limited "success" (one of the subjects could detect relative light levels, but not general shapes, at the cost of severe migraines) As I recall, this research was considered unethical in the US due to the invasive nature of the brain implant related to the expected outcome.

Yep, you're right. I just read through the full text version of one of the papers and the monkeys appear to be able to selectively move between using their real arm, the robotic arm, or both. Maybe the real challenge is making the additional arm(s) highly funtional rather than just getting it to work on a basic level. In game terms, perhaps a combination of something like the cerebral booster implant, but targeted at motor/sensory pathways rather than frontal cortex combined with some of the drug treatments mentioned in the material about cyborgs that allow for a higher degree of neuronal growth. This would insure that you have enough brain "real-estate" for fine manipulation of the hand and coordinating movements with other limbs. You would also need something along the lines of muscle augmentation (or the bioware equivalent) so that the new arm(s) has all of the necessary associated musculature along the trunk (or where ever you attach the thing.)

[Mardrax]

Affixation of cyberlimbs is done proximal to the joint, not in it, or even just the distal end, as tends to be the case nowadays.
If you get a cyber lower leg, the cybering would start probably at the distal epiphyses of the femur. A full arm includes a functional shoulder joint. At least a partial replacement of all musculature in between the chin and the midriff, excepting spinal flex/erectors should probably be part and parcel of this, since really, an arm can't function without that subclavial, trapezium or pectoral. Actually needing to have additional modifications of surrounding structures for support only comes into play for attribute enhancements over 3, in the form of a cybertorso.
The same should hold true for any extra limbs.

And really, any "newly acquired limb" functions on a very low level. An infant can't manipulate its hands beyond simply grasping and extending. In fact, all muscles tend to function like this. Just try and elevate your left eyebrow, keeping the other still. Or raise the right side of your tongue, without lip-aid. You could do both, but it would require more training than the average human has. This development takes place in several places. Notably in the motor cortex, where a cognitive mapping of the body's motor functions is maintained, and in the cerebellum, where the general commands from the motor cortex are modified to control motor units (groups of motor fibres in a single muscle that are controlled by a single neuron) that in turn cause motor fibres to contract. The more developped individual will have both a greater ammount of detail in his motor cortex, a wider ammount of comprehensible "commands" in his cerebellum, and a better tweaked system of motor units.
If you've ever taken up systematical physical exercise, and noticed an improvement in your area of exercise within three weeks or so, that's because your system of motor units has adapted for better control. By way of experiment, you could reverse this process and tape a few fingers together for a few days, then see how hard it is to control them separately afterwards. Broken limb victims will be familiar with this.

I think a monkey that is given a robotic arm with functionality more akin to a natural arm could well learn to use most of its functions with more than passing ability. It would just take a massive ammount of time. Remapping brain functions to new (and "unintended") functionality really is a given by now, exposed amazingly by younger patients who've had a hemispherectomy. If you're interested in that kind of stuff, there's loads of source material available.