Prediction Errors Drive Learning in a Mouse Model of Motor Adaptation
MetadataShow full item record
CitationMathis, Mackenzie. 2017. Prediction Errors Drive Learning in a Mouse Model of Motor Adaptation. Doctoral dissertation, Harvard University, Graduate School of Arts & Sciences.
AbstractThe brain builds internal models of our body and environment. The nature and purpose of these models have been a matter of philosophical and scientific debate for centuries. In neuroscience, seminal behavioral experiments have demonstrated their existence and essential role in successful motor actions; however, where in the brain these models reside, how they are formed, and how they are updated following bodily or environmental changes remain unclear. Today, with the advent of techniques to measure and manipulate hundreds of single cells in a behaving animal, these questions can finally be explored. Here, in Chapter 2 I describe the development of a novel behavioral system to teach mice to perform skilled reaching tasks. In Chapter 3, inspired by primate research, I developed a novel adaptation paradigm for mice where they learn to account for a limb perturbation generated by a force field. Using precisely timed optogenetics, I have found that the primary forelimb somatosensory cortex (S1) is required for adapting to this limb perturbation, and the data suggests S1 may be updating the animal’s internal model of the force field. Lastly, in Chapter 4, I test the role of both reward and sensory perturbations to unveil that limb force field perturbations are guided primarily by sensory prediction errors, but only when they are task relevant. Together, my results provide a mouse model system of adaptation and explores the neural mechanisms underpinning motor adaptation.
Citable link to this pagehttp://nrs.harvard.edu/urn-3:HUL.InstRepos:41142031
- FAS Theses and Dissertations