The Role of Error-sensitivity in Motor Adaptation

When we experience an error in a motor task, we adapt our next movement to partially compensate. The process of adaptation can be modeled as u(n+1)=αu(n)+η(n)e(n) where u(n) is the motor command on trial n, α is a decay factor, e(n) is error, and η(n) represents the subjects’ sensitivity to the experienced error. Here, we explore the rules that govern the value of η(n) as well as the brain-regions that are responsible for its evaluation. In Chapter 2, we begin with a puzzle: in motor learning tasks, humans are able to modulate how much they learn from a given error. In some conditions, they learn a large amount, but in other conditions they learn only a small amount. That is, the brain selects how much it is willing to learn from error. We suggest that ‘error-sensitivity’ is modulated by the history of previous errors. What brain region is responsible for determining the amount subjects are willing to learn from an error? Adaptation is critically dependent on the cerebellum, as demonstrated by patient and lesion studies. In Chapter 3 we use transcranial direct current stimulation (tDCS) to alter the function of the cerebellum, and observe its effects on error-sensitivity. We find that increasing the excitability of the cerebellum via anodal tDCS increases the rate of learning, while decreasing iicerebellum excitability via cathodal tDCS decreases error-sensitivity. That is, we suggest the cerebellum is responsible for determining how much subjects are willing to learn from a motor error. How does the cerebellum accomplish the task of adaptation? It is has been proposed that the firing rates of the principal cells of the cerebellum, Purkinje (P-)cells, should encode movement kinematics. Yet, this has remained a long standing puzzle, as no clear encoding of movement kinematics has been found. How the cerebellum learns has been difficult to approach because the problem of encoding remains unresolved. In Chapter 4 we approach this problem from a new direction: we propose that the cerebellum is composed of micro-clusters of P-cells, organized based on their preference for error. When the cells are organized in this manner, a clear encoding of kinematics emerges