To think or not to think?
By Po-Chen Liu, class of 2020.
Visit Po-Chen’s LinkedIn and Website.
Mind control, is real, well basically.
In 2012, a paper published in the journal Nature by Brown University demonstrated paralysed subjects can control a robotic arm using their thoughts, and only their thoughts, via a Brain Computer Interface (BCI).
A new hope in the future of restoring damaged bodies using robotic prostheses? Cyborgs?!
One of the subjects was Cathy Hutchinson. She’s a stroke survivor who is now a tetraplegic spending her life without the power of speech and a wheelchair user. Something you and I might ask is, what sort of life does she live?
The answer I’m afraid, is just what you’d imagine it to be.
Cathy also has a BCI chip the size of an aspirin (shown above) implanted on the surface of a region in the brain called the motor cortex which is responsible for sending signals to your nervous system and instructing your muscles to contract and relax. These electrical signals are called EEG (encephalogram) signals and can be picked up by electrodes on the chip to determine the power of the signal.
How does a Brain Computer Interface work?
Now EEG signals are incredibly arduous to work with because your brain is complex and amazing!
It’s doing a myriad of other things such as keeping you alive, making sure you blink, thinking-about-an-embarrassing-moment-that-happened-5-years-ago-at-2am, you know. The usual.
This means there are many other signals interfering with the signals related to movement which makes your signals…uh…well see for yourself.
No human can process such obscure information by visually inspecting it, however, we can train an algorithm to recognise certain patterns correlated to moving your arm or your hand. Patterns such as, you ready?
Event Related Synchronisation and Desynchronisation and Bereitschaftspotential are a few of the unique patterns created by motor imagery. Motor imagery is when you only think about moving, but don’t. It still results in EEG signals from your motor cortex but more often than not, is absolutely indistinguishable to the human eye. Here’s a sneak peak at one of those patterns~
And when I say train, I mean quite literally train. A key component in these types of algorithms for BCI is ‘punishing’ the algorithm when it gets something wrong, which informs the algorithm to improve itself. Surely this is a way to get robots to hate us?
Not everyone gets to sip a cinnamon latte out of a robotic hand. For Cathy Hutchinson, it’s not the taste of coffee she cares for, but the first taste of independence she’s had in over 15 years.
The smile on her face was a remarkable thing to see. For all of us involved, we were encouraged that the research is making the kind of progress that we had all hoped
said Leigh Hochberg, an associate professor of engineering at Brown University.
Other applications
Not only does this technology enable function, it can detect anomalies and rehabilitate!
By training the algorithm to recognise ‘correct’ brain signals, you can rehabilitate stroke patients by training them to ‘think correctly’ and send the right signals through a damaged signal channel to relearn how to control a part of your body.
It even helps diagnose epilepsy as there are numerous types of seizures, not just the ‘flailing and flopping’ kind you see in movies but also subtle ones that occur without notice. These can cause serious issues such as stopping in the middle of the road or simply forgetting your name!
Implications of this technology?
Think about, literally. What we have touched on today is the fact we can bypass our physical limitations by directly accessing the source of information, our brain. This impacts everything we have created. Because the fundamental denominator in every single human oriented design you can think of, is thinking.
But personally what do I look forward to?
To see it restore functionality in disabled people that you and I take for granted. For them, to do the things they love, to just live life a little.
That is what excites me about Brain Computer Interfaces, what about you?
Further information
Cathy Hutchinson BrainGate Brain Computer Interface demonstration
Elon Musk’s Brain Computer Interfaces company, Neuralink
Brain Computer Interface with EEG signals – moderate details
For experts who want to delve deeper into the details of motor imagery