Instead of just trying out various firing patterns on the paralysed muscles in the hope of finding one that works, you can use computer models of the musculoskeletal system to calculate them. These models are mathematical descriptions of how muscles, bones and joints act and interact during movement. In the simulations, you can make muscles stronger or weaker, “paralysed” or “externally stimulated”. You can test different firing patterns quickly and safely, and you can make the models pick up their tea cups over and over again – sometimes more successfully than others.
Modelling The Muscles
To test the technology, the team at Keele is working with the Cleveland FES Center in the US, where they implant up to 24 electrodes into the muscles and nerves of research participants. They use modelling to decide where to place the electrodes because there are more paralysed muscles than electrodes in current FES systems.
If you have to choose, is it better to stimulate the subscapularis or the supraspinatus? If you stimulate the axillary nerve, should you place the electrode before or after the branch to the teres minor? To answer these difficult questions, they run simulations with different sets of electrodes and choose the one that allows the computer models to make the most effective movements.
Currently, the team is working on the shoulder, which is stabilised by a group of muscles called the rotator cuff. If you get the firing instructions for the arm wrong, it might reach for the soup spoon instead of the butter knife. If you get the instructions to the rotator cuff wrong, the arm might pop out of the shoulder. It is not a good look for the computer models, but they don’t complain. Research participants would be less forgiving.
Knowing how to activate paralysed muscles to produce useful movements like grasping is only half of the problem. We also need to know when to activate the muscles, for example when the user wants to pick up an object. One possibility is to read this information directly from the brain. Recently, researchers in the US used an implant to listen to individual cells in the brain of a paralysed individual. Because different movements are associated with different patterns of brain activity, the participant was able to select one of six pre-programmed movements that were then generated by stimulation of hand muscles.