Sensorimotor learning is an essential feature of human competence, allowing people to flexibly move in novel and variable environments. The cerebellum is a critical structure in sensorimotor pathways as evidenced by the fact that deficits in sensorimotor learning and control are the prominent features of ataxia, a neurological condition observed in individuals who suffer from degenerative disorders of the cerebellum. The over-arching goal of this project is to advance our understanding of the role of the cerebellum in sensorimotor learning. Recent experimental work has shown that sensorimotor learning entails the operation of multiple learning processes, some of which are under volitional control and others that operate in an automatic and implicit fashion. Identifying the contribution of these different processes has been hindered by the use of experimental methods that confound information arising from different feedback signals. The key experimental manipulation to be used in the proposed research will entail a new method, designed to isolate cerebellar-dependent error based adaptation, a learning process that serves to ensure the calibration of sensorimotor maps. This method involves a manipulation of visual feedback such that the error signal remains invariant despite changes in behavior. Preliminary results with this method reveal fundamental inadequacies with current computational models of sensorimotor adaptation. To address these limitations, the proposed work entails the integrated use of behavioral, neuroimaging, and computational methods to develop new models of sensorimotor adaptation. The research plan centers on three specific aims.
The first aim i s to iteratively use computational and behavioral methods to develop and test new computational models of sensorimotor adaptation. Psychophysical experiments with healthy young adults will be conducted to specify the constraints on error-based adaptation, with this information used to inform model refinement. The goal of the second aim is to characterize the contribution of the cerebellum to error-based learning, testing key predictions from the modeling work concerning the sensitivity of this structure to error information. This work will entail behavioral studies in patients with cerebellar degeneration and, as a point of contrast, patients with Parkinson's disease, as well as functional neuroimaging (fMRI) studies in healthy young adults. The goal of the third aim is to examine constraints on the error signal that drive adaptation, and in particular, explore how adaptation may interact with systems sensitive to task outcome. Various manipulations will be employed that may modulate the the strength of error information, or put this information in conflict with reinforcement signals. Taken together, this project should provide a new framework for understanding the contributions of the cerebellum to sensorimotor adaptation and gain insight into the behavioral changes observed in individuals with ataxia arising from cerebellar degeneration.
Sensorimotor learning is an essential feature of human competence, allowing people to flexibly move in novel and variable environments. This ability arises from the combined operation of multiple learning processes, some of which are under volitional control and others that operate in an automatic and implicit fashion. The focus of this project is on error-based learning that is dependent on the cerebellum, using this information to refine computational models of how this subcortical system contributes to motor learning and better understand the behavioral changes observed in individuals with ataxia arising from cerebellar degeneration.
| Breska, Assaf; Ivry, Richard B (2018) Double dissociation of single-interval and rhythmic temporal prediction in cerebellar degeneration and Parkinson's disease. Proc Natl Acad Sci U S A 115:12283-12288 |
