Investigating the role of representational capacity in sensorimotor learning with MRI neuroimaging

Abstract

This thesis explores sensorimotor learning (SL) in the brain, focusing on its efficiency, generalization ability, and resilience to interference. It delves into how the brain's representational capacity (RC) - the number of possible neural representations - influences such aspects of SL. Our theoretical description suggests a direct relationship between RC and learning speed: faster learning tends to occur when the RC is smaller in the information-sending region and larger in the receiving region. Moreover, the description proposes that RC also impacts generalization and interference. MRI experimental findings reveal associations between gray matter volume (GMV) and learning rates. Initial performance is positively correlated with cerebellar GMV and negatively with dorsomedial prefrontal cortex (DMPFC) GMV. Early learning rate shows a positive correlation with dorsolateral prefrontal cortex (DLPFC) GMV and a negative correlation with cerebellar GMV. Late learning rate correlates positively with thalamic GMV and negatively with cerebellar GMV. These correlations suggest the cerebellum's vital role across different adaptation stages, the prefrontal involvement in early learning, and the thalamus's role in later stages. Functional data further supports those involvements. Additionally, the study finds that conditions favoring rapid learning hinder motor generalization. MRI data shows a negative correlation between generalization and GMV in the cerebellum and putamen, with positive correlations in cortical areas. Interference is negatively correlated with GMV in the caudate nucleus and cerebellum. The theoretical description of RC provides a concise explanation of these intricate findings. In conclusion, this research advances understanding of the impact of GMV on key aspects of human SL and underscores the need for further investigation into the GMV-RC association. The current findings highlight the comprehensive role of the cerebellum, the significance of the prefrontal cortex in early learning, and the contribution of the caudate nucleus in skill re-acquisition, establishing a foundation for understanding the brain's SL capabilities.