Electric muscles
In an earlier post, I mentioned that we can use the contractions of the hand muscles located in the forearm to control a prosthetic hand. But how do we detect that a muscle is contracting?
Here’s where a technique called electromyography (EMG for short) comes in. It has a long and unappealing name, but when I demonstrate it in science festivals, it is always a big hit!
Imagine the following scenario. There is a chocolate biscuit on the table in front of you. You want the biscuit. So you move your hand and take it. How did you do that?
When you decide to move your hand, the motor cortex, which is the part of your brain responsible for movement, sends electrical signals down the nerves to the spinal cord, where they connect to other nerves that carry these electrical signals all the way to the appropriate hand muscles.
(In case you are wondering, these electrical signals travel really fast, at around 50 metres per second. So they can get from your shoulder to your forearm in less than one hundredth of a second! That’s why you don’t notice any delays between thinking of picking up the biscuit and actually doing it.)
When the electrical signals reach the muscles, they cause them to contract. The muscles are attached to bones of the forearm and hand, so when they contract they pull on the bones. (Based on the exact configuration of muscles and bones, the mechanics can get quite complex! More on that in a future post.) Your hand moves to the biscuit, and the rest is history.
With EMG, we can record the electrical activity of the muscles as they contract. We do this with sensors that we stick on the skin, on top of the muscle we are interested in. Kids in science festivals love it when they realise their body is electric!
Prostheses that use EMG for control are called myoelectric. The prosthesis user contracts muscles in the forearm that are associated with specific functions of the prosthetic hand, for example a certain way of grasping. EMG sensors in the prosthesis socket record the electrical activity of these muscles, and as a result the prosthetic hand performs the desired grasp.
At the moment, myoelectric prosthetic hands are not capable of moving in ways that are not predetermined (such as performing certain grasps), because we record from a small number of muscles, and associate their electrical activity with specific functions. If we recorded from more muscles, could we get prosthetic hands to be more spontaneous?
