5 ms·

Walking naturally after spinal cord injury using a brain–spine interface

338 points by Someone 3y ago

phobotics 3y ago
I’m very excited to see how this progresses and how A.I. can help solve medical issues that have up to now stumped us. One really interesting thing they’ve reported about this is that it requires a head set to connect the brain implants to the spine electrodes, but after a while the patient was able to walk using a stroller without the headset because the body regrew nerve tissue and Reenabled some connection to the brain.
- padolsey 3y ago
  It's very cool, and uses the same principles as lots of other neurorehab electrical devices out there. I used an FES ('Functional Electrical Stimulation') device for a while that has a little accelerometer in it to figure out when I intended to take a step, and then stimulate (via conductive pads placed atop the skin) the correct muscles in the lower leg to lift the ankle up (dorsiflexion). Over time, when the brain correlates X input with Y result, even if Y is very very weak, bruised nerves can recover (in the case of recent brain/spinal injuries) and new nerves can grow. I was advised to have the device on its 'training' setting (where it would periodically activate a 'step') even when lounging on the sofa. Hebb said: "What fires together, wires together"... My suspicion (tho I'd love to be educated on this) is that the reference patient did not have a truly 'complete' spinal cord injury, but rather one where there were at least some nervecells remaining/bruised. If there was a complete section of nerve wiring missing, I don't see how re-growth could occur (but perhaps there's some stem-cell wizardry that can help with that???). Anyway, it wouldn't matter much; if these devices become small and convenient enough, it's not much burden to wear them forever.
  - Gh0stRAT 3y ago
    Yeah, if you watch the supplemental videos they mention that the patient had a partial injury and after 7 years of physical therapy was still unable to walk independently. After getting this implant and doing 6 months of PT, the patient was able to walk unassisted even with the brain interface not powered on.
  - bmogen 3y ago
    Yes, complete spinal cord injuries (think 'total cut') are pretty rare compared to incomplete injuries (think 'crush') [1] https://www.nscisc.uab.edu/Public/Facts%20and%20Figures%202019%20-%20Final.pdf https://www.nscisc.uab.edu/Public/Facts%20and%20Figures%2020...
  - LegionMammal978 3y ago
    > My suspicion (tho I'd love to be educated on this) is that the reference patient did not have a truly 'complete' spinal cord injury, but rather one where there were at least some nervecells remaining/bruised. This is mentioned in the article at the start of the Discussion section: > The validation of this digital bridge was restricted to a single individual with severe but partial damage of the spinal cord, and it therefore remains unclear whether the BSI will be applicable to other injury locations and severities. However, several observations suggest that this approach will be applicable to a broad population of individuals with paralysis. The authors likewise anticipate that there won't be any technical hurdles in miniaturizing the technology and making it self-calibrating.
  - sheepscreek 3y ago
    > ..the most notable achievement is the fact that these three patients had complete spinal cord injury, instead of a partial one as with the 2018 patients. Patients with “complete” spinal cord injury have been successfully treated as well: https://english.elpais.com/science-tech/2022-02-08/three-paralyzed-patients-walk-again-thanks-to-spinal-cord-implants.html#?rel=mas https://english.elpais.com/science-tech/2022-02-08/three-par...
aeturnum 3y ago
This is very exciting! One possibility that is, I think, covered up by all the hype around "AI works like our brains" (which I disagree with) is that: our brains are too complex for us to understand and ML may allow us to build more capable interfaces that we still don't understand - but which are easier to understand than the brain. It's an extremely promising new avenue for incremental progress.
- carlmr 3y ago
  >"AI works like our brains" (which I disagree with) While AI works like our brains is patently false it might just be analogous enough to not matter in the end if the details are the same to get similar emergent behavior.
  - cookieperson 3y ago
    Most applications like these don't require "emergent behavior" at all really... It's basically domain mapping from data. Super cool, but it's not like these models are all of a sudden forming personalities or something...
  - tltimeline2 3y ago
    Insert "Two black boxes walk into a bar" joke here.
- ElfinTrousers 3y ago
  "AI works like our brains" is indeed generally not so, but CNNs do work like the visual cortex, somewhat--it's where they got the idea for them in the first place.
0xbadc0de5 3y ago
Always good to see novel uses of ML. However, let's not pretend this will ever offer any benefit to people with spinal cord injuries. Any research that does not involve repairing the physical damage to the spinal cord itself is a distraction. Anyone who claims otherwise does not understand the physiology involved or is being deliberately disingenuous. It's hard to think of an apt analogy, but perhaps consider a patient who walks into a doctor's office with a broken arm. Instead of setting the break and applying a cast, the doctor offers them a referral to a psychologist to help them adjust to life with a broken arm.
- _joel 3y ago
  > Always good to see novel uses of ML. However, let's not pretend this will ever offer any benefit to people with spinal cord injuries. Apart from the person in their study who can now walk, that had the spinal injury (with the use of a crawler)... In relation to your second point, you are aware regenereative science and this implant approach can progress in parallel? This offers something tangible, plus with more refining can be rolled out en masse. What regenerative science can be applied now that would give such a quality of life improvement?
  - 0xbadc0de5 3y ago
    The research subject has an incomplete C5/C6 injury and still cannot walk unassisted. Not to mention still has limited use of their hands and likely severely impacted somatosensation. It's not physically possible to manufacture implants dense enough to re-map the entirety (or even a useful fraction) of motor and sensory pathways. Attempting to implant a significant number of artificial electrodes also risks further damage. I'm not claiming that regenerative treatments and implants can't progress in parallel. I'm simply claiming implants are a dead-end and are distracting resources away from regenerative research.
    - _joel 3y ago
      You literally said this wouldn't help spinal injuries, which the PoC has and now you're changing it to 'unassisted' Do you have data to show funding gets moved away from regenerative research?
_joel 3y ago
Saw this on the news just now https://www.bbc.co.uk/news/science-environment-65689580 https://www.bbc.co.uk/news/science-environment-65689580
barelyauser 3y ago
What is the practical reason to favor electrical implants over electromagnetic ones? One would conclude that using electromagnetic induction would be a no-brainer when it is very clear the body works against electrical probes (by fighting the implants). Could someone more knowledgeable than myself clarify this.
- jupp0r 3y ago
  I don't know the answer to your question, but I appreciate the "no-brainer" pun there.
- 0xbadc0de5 3y ago
  It's very difficult (ie: impossible) to target individual nerve fibres with induction. Direct stimulation is the only way to achieve the required resolution.
  - hprotagonist 3y ago
    see also, MEG
- etrautmann 3y ago
  The amount of current required for magnetic stimulation is much higher and not compatible with implantable devices. TMS requires large coils and currents relative to electrical stimulation.
  - barelyauser 3y ago
    Isn't TMS coil current high to compensate the distance? The closer to the nerve ending should reduce the necessary current.
    - Jeff_Brown 3y ago
      But if those nerve endings are anywhere but at the surface of the body, they'll still require an implant.
belter 3y ago
"A quadriplegic man walks again, thanks to a digital bridge between his brain and spinal cord" - https://english.elpais.com/science-tech/2023-05-24/a-quadriplegic-man-walks-again-thanks-to-a-digital-bridge-between-his-brain-and-spinal-cord.html https://english.elpais.com/science-tech/2023-05-24/a-quadrip...
__loam 3y ago
This is genuinely incredible.
verisimi 3y ago
Right, another story to help us on to our knees into a begging position.... "Please give us a neural lace!" Says no one... yet. But the priming is underway!
- ninjanomnom 3y ago
  Speak for yourself there, I've been in favor of brain augments essentially ever since I was aware of them as a concept as a teen and hoping they'd arrive in my lifetime. Personally it's because I don't hold most of my identity as sacred and would like to have full control over my automatic behaviors and thought patterns.
  - verisimi 3y ago
    Yes. People like Iain M Banks and other scifi makes a good sales pitch. But what about the opportunity to live a natural life, attempting to find the full expression as God (whatever that is) created us? And do you really feel comfortable giving a tech company access to your mind?
dr_ 3y ago
As a student I was taught that some day stem cells would allow sci patients to recover. Instead, hard tech seems to be the more apparent and workable solution.

338 points by Someone 3y ago

I’m very excited to see how this progresses and how A.I. can help solve medical issues that have up to now stumped us. One really interesting thing they’ve reported about this is that it requires a head set to connect the brain implants to the spine electrodes, but after a while the patient was able to walk using a stroller without the headset because the body regrew nerve tissue and Reenabled some connection to the brain.

padolsey 3y ago
It's very cool, and uses the same principles as lots of other neurorehab electrical devices out there. I used an FES ('Functional Electrical Stimulation') device for a while that has a little accelerometer in it to figure out when I intended to take a step, and then stimulate (via conductive pads placed atop the skin) the correct muscles in the lower leg to lift the ankle up (dorsiflexion). Over time, when the brain correlates X input with Y result, even if Y is very very weak, bruised nerves can recover (in the case of recent brain/spinal injuries) and new nerves can grow. I was advised to have the device on its 'training' setting (where it would periodically activate a 'step') even when lounging on the sofa. Hebb said: "What fires together, wires together"... My suspicion (tho I'd love to be educated on this) is that the reference patient did not have a truly 'complete' spinal cord injury, but rather one where there were at least some nervecells remaining/bruised. If there was a complete section of nerve wiring missing, I don't see how re-growth could occur (but perhaps there's some stem-cell wizardry that can help with that???). Anyway, it wouldn't matter much; if these devices become small and convenient enough, it's not much burden to wear them forever.
- Gh0stRAT 3y ago
  Yeah, if you watch the supplemental videos they mention that the patient had a partial injury and after 7 years of physical therapy was still unable to walk independently. After getting this implant and doing 6 months of PT, the patient was able to walk unassisted even with the brain interface not powered on.
- bmogen 3y ago
  Yes, complete spinal cord injuries (think 'total cut') are pretty rare compared to incomplete injuries (think 'crush') [1] https://www.nscisc.uab.edu/Public/Facts%20and%20Figures%202019%20-%20Final.pdf https://www.nscisc.uab.edu/Public/Facts%20and%20Figures%2020...
- LegionMammal978 3y ago
  > My suspicion (tho I'd love to be educated on this) is that the reference patient did not have a truly 'complete' spinal cord injury, but rather one where there were at least some nervecells remaining/bruised. This is mentioned in the article at the start of the Discussion section: > The validation of this digital bridge was restricted to a single individual with severe but partial damage of the spinal cord, and it therefore remains unclear whether the BSI will be applicable to other injury locations and severities. However, several observations suggest that this approach will be applicable to a broad population of individuals with paralysis. The authors likewise anticipate that there won't be any technical hurdles in miniaturizing the technology and making it self-calibrating.
- sheepscreek 3y ago
  > ..the most notable achievement is the fact that these three patients had complete spinal cord injury, instead of a partial one as with the 2018 patients. Patients with “complete” spinal cord injury have been successfully treated as well: https://english.elpais.com/science-tech/2022-02-08/three-paralyzed-patients-walk-again-thanks-to-spinal-cord-implants.html#?rel=mas https://english.elpais.com/science-tech/2022-02-08/three-par...

aeturnum 3y ago

This is very exciting! One possibility that is, I think, covered up by all the hype around "AI works like our brains" (which I disagree with) is that: our brains are too complex for us to understand and ML may allow us to build more capable interfaces that we still don't understand - but which are easier to understand than the brain. It's an extremely promising new avenue for incremental progress.

carlmr 3y ago
>"AI works like our brains" (which I disagree with) While AI works like our brains is patently false it might just be analogous enough to not matter in the end if the details are the same to get similar emergent behavior.
- cookieperson 3y ago
  Most applications like these don't require "emergent behavior" at all really... It's basically domain mapping from data. Super cool, but it's not like these models are all of a sudden forming personalities or something...
- tltimeline2 3y ago
  Insert "Two black boxes walk into a bar" joke here.
ElfinTrousers 3y ago
"AI works like our brains" is indeed generally not so, but CNNs do work like the visual cortex, somewhat--it's where they got the idea for them in the first place.

0xbadc0de5 3y ago

Always good to see novel uses of ML. However, let's not pretend this will ever offer any benefit to people with spinal cord injuries. Any research that does not involve repairing the physical damage to the spinal cord itself is a distraction. Anyone who claims otherwise does not understand the physiology involved or is being deliberately disingenuous. It's hard to think of an apt analogy, but perhaps consider a patient who walks into a doctor's office with a broken arm. Instead of setting the break and applying a cast, the doctor offers them a referral to a psychologist to help them adjust to life with a broken arm.

_joel 3y ago
> Always good to see novel uses of ML. However, let's not pretend this will ever offer any benefit to people with spinal cord injuries. Apart from the person in their study who can now walk, that had the spinal injury (with the use of a crawler)... In relation to your second point, you are aware regenereative science and this implant approach can progress in parallel? This offers something tangible, plus with more refining can be rolled out en masse. What regenerative science can be applied now that would give such a quality of life improvement?
- 0xbadc0de5 3y ago
  The research subject has an incomplete C5/C6 injury and still cannot walk unassisted. Not to mention still has limited use of their hands and likely severely impacted somatosensation. It's not physically possible to manufacture implants dense enough to re-map the entirety (or even a useful fraction) of motor and sensory pathways. Attempting to implant a significant number of artificial electrodes also risks further damage. I'm not claiming that regenerative treatments and implants can't progress in parallel. I'm simply claiming implants are a dead-end and are distracting resources away from regenerative research.
  - _joel 3y ago
    You literally said this wouldn't help spinal injuries, which the PoC has and now you're changing it to 'unassisted' Do you have data to show funding gets moved away from regenerative research?

_joel 3y ago

Saw this on the news just now https://www.bbc.co.uk/news/science-environment-65689580 https://www.bbc.co.uk/news/science-environment-65689580

barelyauser 3y ago

What is the practical reason to favor electrical implants over electromagnetic ones? One would conclude that using electromagnetic induction would be a no-brainer when it is very clear the body works against electrical probes (by fighting the implants). Could someone more knowledgeable than myself clarify this.

jupp0r 3y ago
I don't know the answer to your question, but I appreciate the "no-brainer" pun there.
0xbadc0de5 3y ago
It's very difficult (ie: impossible) to target individual nerve fibres with induction. Direct stimulation is the only way to achieve the required resolution.
- hprotagonist 3y ago
  see also, MEG
etrautmann 3y ago
The amount of current required for magnetic stimulation is much higher and not compatible with implantable devices. TMS requires large coils and currents relative to electrical stimulation.
- barelyauser 3y ago
  Isn't TMS coil current high to compensate the distance? The closer to the nerve ending should reduce the necessary current.
  - Jeff_Brown 3y ago
    But if those nerve endings are anywhere but at the surface of the body, they'll still require an implant.

belter 3y ago

"A quadriplegic man walks again, thanks to a digital bridge between his brain and spinal cord" - https://english.elpais.com/science-tech/2023-05-24/a-quadriplegic-man-walks-again-thanks-to-a-digital-bridge-between-his-brain-and-spinal-cord.html https://english.elpais.com/science-tech/2023-05-24/a-quadrip...

__loam 3y ago

This is genuinely incredible.

verisimi 3y ago

Right, another story to help us on to our knees into a begging position.... "Please give us a neural lace!" Says no one... yet. But the priming is underway!

ninjanomnom 3y ago
Speak for yourself there, I've been in favor of brain augments essentially ever since I was aware of them as a concept as a teen and hoping they'd arrive in my lifetime. Personally it's because I don't hold most of my identity as sacred and would like to have full control over my automatic behaviors and thought patterns.
- verisimi 3y ago
  Yes. People like Iain M Banks and other scifi makes a good sales pitch. But what about the opportunity to live a natural life, attempting to find the full expression as God (whatever that is) created us? And do you really feel comfortable giving a tech company access to your mind?

dr_ 3y ago

As a student I was taught that some day stem cells would allow sci patients to recover. Instead, hard tech seems to be the more apparent and workable solution.