There are numerous ways to reach for an apple hanging from a tree. Yet, our motor system uses a specific muscle activity pattern to generate reaching movements that have similar characteristics. For many decades, we know that this pattern features activity bursts and silent periods. We suggest that these bursts are a strong evidence against the common view that the brain continuously controls the commands to the muscles. Instead, we suggest a model that changes these commands in a discrete way. We use unsupervised machine learning to identify transitions in the state of the muscles, and show that fitting a discrete model to the kinematics of movement using only one parameter predicts the transitions in the state of the muscles. Such discrete controller suggests that the brain reduces the complexity of the motor control problem as well as the wear-and-tear of the muscles by sending commands to the muscles at sparse times. Identifying this discrete controller can be applied in the control of prostheses and physical human-robot interaction systems such as exoskeletons and assistive devices.
|Publisher||Cold Spring Harbor Laboratory Press|