| Humans are capable of a wide repertoire of movements and are able to learn new movements or modify previously learned ones when presented with novel mechanical forces. Sensory feedback plays a role in motor learning, yet it is not clear how one associates changes in motor patterns with sensory feedback of motor performance. Neural circuitry for volitional motor control has a natural tendency to create local associations where the sensory and motor fields of individual neurons overlap. Our main hypothesis is that this local sensory-to-motor association creates a default strategy for motor learning. To address this issue, we first developed a novel robotic device for human use which allowed us to measure and manipulate motion of the shoulder, elbow or both during planar multi-joint reaching. In Chapter 3, we describe a study using this device where subjects were trained to compensate for a novel load applied to the elbow during a multi-joint movement where many variables of movement were highly correlated with the load. We show that the motor system assumed the load was related to elbow motion, even when the load was dependent on shoulder motion. This default strategy may reflect the underlying organization of the neural circuitry for motor control, where the overlap of sensory and motor fields of individual neurons creates a local sensory-motor association. We began to investigate this relationship further by reporting a detailed analysis of primary motor cortex neural responses to passive manipulation of the elbow and shoulder. Our results show that many of the cells were responsive to motion of both the shoulder and the elbow, and that a large proportion of these cells tended to be sensitive to the same direction of motion, flexion or extension, at both joints. Previous work in this lab will allow us to directly compare the responses of MI neurons during sensory and motor tasks. Preliminary findings suggest a role for context-dependent modulation of sensory responses. |
PDF Full Text Request