Data collection part 1
I’m writing this on a train going from Newcastle-under-Lyme to Newcastle-Upon-Tyne. We’ll spend the next two and a half days there collecting data for our experiment! My brain is so busy running through checklists and protocols that I can’t even think of a joke to make about the two Newcastles.
Why have we given ourselves only two and a half days? As I mentioned before, doing research comes with constraints in time and resources. In this case, most of the equipment we are using belongs to our collaborators in Newcastle University, and because of teaching and other commitments we could only afford three days away from our university. But I’m confident we can do it!
Let me briefly explain what this experiment is about. It is part of a larger project that aims to use computer modelling to improve the way prosthetic hands move and feel.
When it comes to intact human hands, movement and sensation are effortless, thanks to the excellent communication between our hands and our brains: when we want to move, electrical signals travel from the brain, down the nerves, to the hand muscles that contract and make the hand move. (Remember the chocolate biscuit?) At the same time, information about the position and movement of the hand, which comes from sensors located inside the hand muscles, is sent from the hand back to the brain, also using electrical signals. This is the sense of “proprioception”.
Current prosthetic hands do not have this two-way communication system. A few of the electrical signals from the brain to the muscles are used to open or close the hand in a predefined set of grasp patterns. But most of the information sent to the muscles is not used, and there is no information sent from the prosthetic hand back to the brain.
Our computer hand model can facilitate this communication. If we record all the electrical signals coming from the brain (that would have been used to move the intact hand), we can send them to our computer model, the model will “move”* accordingly, and we can get the hand prosthesis to copy this movement. At the same time, the muscles of the computer model will generate information about the position and movement of the hand, which we can send back to the brain, to give the prosthesis user the sense of “proprioception”.
It sounds pretty good, doesn’t it? We have already shown that our computer model can run fast enough, by which we mean it can solve the necessary equations fast enough to enable the communication to the prosthesis with no delays. What we want to do now is connect it to all the necessary hardware and try it out with actual people!
But it’s a bit more complicated than that. Currently, we don’t have the technology to record all the electrical signals going from the brain to the muscles, or send proprioceptive signals back to the brain. (But we will, other researchers are working on it!) So how can we try out our model?
We’ve had endless discussions about this. If our model had access to all the electrical signals going from the brain to the muscles, it would enable a prosthetic hand to make any movement, not just simple grasps. But since at the moment we can only record from a limited number of muscles, we cannot demonstrate the full capabilities of the model…
At the same time, the model can generate proprioceptive signals, but we cannot use them yet, so we cannot show they can in fact give a user the sense of proprioception.
We’ve decided to do the experiment anyway, because we want to show that our model can work as part of a complete system that uses EMG from hand muscles to enable a prosthetic hand to move in ways that are not predefined, and generate electrical signals that could be used to give a sense of proprioception. We just have to be realistic about the movements we can achieve with a limited number of EMG sensors.
Will we get any useful EMG signals? Will all the hardware and software talk nicely to each other? Will it work at all?? Stay tuned for a follow-up post early next week.
* This is in quotes because the computer model doesn’t actually move, it’s just a bunch of equations!
