Amputee controls prosthetic arms by thinking about them [video] | Hacker News Reader

Amputee controls prosthetic arms by thinking about them [video]

m.youtube.com

74 pointstaf223 comments

Summary of the technique:

1. Surgically reconfigure amputees nerves that once controlled the arm and the hand so they are compatible with their pattern recognition scheme

2. Teach the pattern recognition system to respond to patients neural stimuli to said nerves : “We use pattern recognition algorithms to identify individual muscles that are contracting, how well they communicate with each other, and their amplitude and frequency,” Chi explained. “We take that information and translate that into actual movements within a prosthetic.”

3. Attach the robotic hands to the patient.

Despite the cool robotics and brain control (nerves to be exact), seeing that guy do the one practical thing, picking up a glass and trying to bring it to his lisp after 40 years, was actually the coolest thing.

Amazing that 15 years ago I was watching primates do these things and more with neural implants. Perfect instance where the idea is amazing and revolutionary but the reality is so hard it takes decades if not longer to accomplish.

The reality is if we dumped as much tax money on it as we dump on spaceships landing on comets, we'd be all done with this in 1 year and headlines like this would be old news.

What would be the point of dumping tax dollars into amputee research?

We invest in space because we think it is worthwhile, and still mostly unprofitable (that's changing).

That is not the case with prosthetics. It makes sense to let the market decide how much to invest in this research acoording to the size of the amputee prosthetic market. It frees up the government to work I. Other things and gets rid of the HUGE inefficiencies when it comes to the government trying to accomplish anything.

> What would be the point of dumping tax dollars into amputee research?

Helping real people that need help now.

> We invest in space because we think it is worthwhile, and still mostly unprofitable (that's changing).

That doesn't help real people that need help now.

Remote-controlled robots on comets, versus more affordable and better prosthetics for all. If the whole country could vote I doubt they'd choose the former over the latter. So who is "we" in "We invest in space because we think it is worthwhile"?

I'll turn your next line around:

It makes sense to let the market decide how much to invest in this research acoording to the size of the space exploration and satellite launching market. It frees up the government to work on other things and gets rid of the HUGE inefficiencies when it comes to the government trying to accomplish anything.

I like how you say "and still mostly unprofitable (that's changing)." You're deluding yourself if you think new industries go through a "not yet profitable" stage where they need help from the government and then afterwards they can be set free.

Many amputees are (ex) military. The government probably should fund cost of treating amputees if they lose limbs in service of their country.

You're right - NASA gets way too much money! Almost 3% of what the defense budget is!

What the hell, they should split their budget with research for prosthetics - dying a cyborg on Earth is way better than visiting some rock in space. /s

Why doesn't NASA grow up from the government's training wheels and becomes a for-profit offering its satellite placing services like everyone else is doing to turn up a buck? Why do I have to subsidize this when they could be making their own money?

If 3% of the defense budget is very little then give it to my industry so I can work without having to deliver any tangible product just like the folks at NASA. If they can deliver they can sell it. If no one wants to buy it then we should not be funding an enterprise that produces goods and services that no one wants.

I do think turning the disabled into cyborgs would be a much better investment of money than placing a glorified remote-controller toy car in a rock far away from Earth. Talk about inconsequential!

But yes you're right, NASA goofing off with our money might pay off 100 years from now. That's much better than being able to replace my limbs today if I get into an accident. /SARCASM

A bit of a late reply, but NASA (established by the US government) is not allowed to turn a profit. Probably so that it does sell innovative high tech to the highest bidder.

Not true, real biological impediments that take time to overcome.

Holy crap, after 40 years without hands he can pick objects and manipulate them with relative dexterity.

I wonder could the hands be scaled up to a point where he could drive a "man-dozer" and be a one man construction company :)

It must be incredibly hard to remember how to even use your arms 40 years after not having them. I'm not a neuroscientist, but I would expect that sooner or later the brain would start to rewrite the paths devoted to arm movement to something else. Looks like this guy made a ton of progress though. Very encouraging to see.

Ok, that was fucking cool.

This is amazing.

There is no direct connection to his brain, so he may not b controlling the arms by just 'thinking'. It looks like almost using smaller muscles to control the arms, but using the nerve connections themselves. Also, he controls one pivot per action - so he moves the shoulder, then the forearm, then the hand, then the fingers with rests between each step.

This is the one thing I didn't get: if it's mapped to nerves why can't he do two pivots at a time? Is it a limitation of the robotics?

It's mapped to the base of nerves. He doesn't have the entire nervous network that an arm usually has.

Here is what I keep thinking when I see these prosthetics.

After you train to control the arm while it's attached, can you more it around regardless of it being actually connected to your shoulder (assuming of course the sensors stay in place) and move it around like thing.

Must not have watched the video:

https://www.youtube.com/watch?v=9NOncx2jU0Q#t=250

This part of the video shows the arms are not attached. I'm sure he was trained to use the arms when they were not on him first.

I am excited for these kinds of technology for non-amputee use too: https://www.thalmic.com/en/myo/

Received a myo yesterday. Pretty awesome hardware execution - aka it works consistently - but the "things" you can do out-of-the-box are fairly limited. I will dig into the SDK over the weekend.

My lifelong dream of becoming Dr. Octopus is one step closer :)

One (close) day such things will be in fashion magazines

Reply on news.ycombinator.com