First mind-controlled leg prothesis
September 10, 2012
BCI-RoGO system (credit: Long Beach Veterans Affairs Medical Center)
Researchers at the Long Beach Veterans Affairs Medical Center in California have built and tested the first prosthetic lower limb that can be controlled in real time by EEG (electroencephalogram) signals.
An EEG signal is fed to brain-computer interface (BCI) computer, which then controls a robotic gait orthosis (RoGO), which stimulates leg muscles.
The experimental setup showed the subject suspended in the RoGO, while donning an EEG cap, surface EMG electrodes, and a gyroscope on the left leg. A monitor (not shown), placed in front of the subject at eye-level, presented instructional cues.
It may be the first step towards developing a biomechanical means to restore able-bodied-like ambulation after spinal cord injury.
In previous work, they developed a way of using EEG signals to control the walking motion of an avatar in a virtual environment.
Future work is necessary to test this system in individuals with spinal cord injury paraplegia. Since spinal cord injuryusers are able to operate the BCI-walking simulator, it is expected that they can readily transfer their skills to the BCI-RoGO system.
Comments (6)
by dagautier
Thanks for this very interesting post, I used it in one of mine about prosthetic legs.
http://dagautier.tumblr.com/post/35304983670/reparer-augmenter-hacker
by SyntaxHD
So in other words, they bypassed the spinal cord as transmission device and sent the signals straight down to the leg muscles via EEG brain-reading technics? Why is it necessary to stimulate the muscles at all, when the brainsignals can be directly transported to the leg robot device?
Have I missunderstood something here?
by Editor
Good question. It would be ideal to pick up the signals that the brain sends to the spinal cord and use those signals directly to control the RoGo (robotic gait orthosis) device (similar to wiring around the fault in an electrical circuit). But as I understand it, that would involve highly invasive (and dangerous) surgery to tap into either the cerebellum (which controls muscles, and is buried deep in the brain, so there’s no way to pick up its signals directly) or higher up on the spinal cord, and such a procedure would confine the patient to a hospital bed in a research lab — that sorta defeats the purpose of enabling them to walk :) (http://www.csuchico.edu/~pmccaffrey/syllabi/CMSD%20320/362unit7.html has a good diagram that makes this clear).
Also, I don’t know of any research in using such signals for RoGo (but I could be wrong about that).
So instead, the non-invasive BCI-RoGo system converts the EEG signal via signal processing and pattern recognition algorithms (effectively simulating what happens in the cerebellum and spinal cord) into a simulated neural signal, which is sent to the muscle. Does that make sense?
(I welcome any corrections from neuroscientists or physicians.)
by melajara
Indeed, this makes perfectly sense. And don’t forget the contribution of the peripheral nervous system to relay and aggregate nerve signalling, this is also superseded by the RoGo here.
Now I much prefer the organic repair approach, backed by stem cells research, and innovative biotech startups, like e.g. http://www.stemcellsinc.com/
by Editor
John, this particular work was restricted to spinal cord injury.
by john lynch
This should help in mind control of vascular Parkinson’s