The use of motor abundance in controlling coordinated bimanual drawing tasks.

摘要

The arm has more muscles and joints, referred to as degrees of freedom (DOFs), than are strictly necessary to achieve the goal of a functional task. How does the nervous system manage these abundant DOFs has become a recent interest in the study of movement control. This dissertation sought to understand inter-limb and intra-limb differences in the organization of abundant DOFs and how these effects were affected by movement dynamics when performing a bimanual rhythmical drawing task. Ten healthy, right-handed subjects participated in each of the three studies using four upper extremity joints. Coordination of these joints was addressed using the framework of the uncontrolled manifold hypothesis (UCM), which allows partitioning of joint variance into two components in relation to the control of task-related variables, such as the hand's drawing trajectory. One component of joint variance represents equivalent patterns of coordination that are consistent with a stable value of, for example, one hand's movement path (goal-equivalent variance, GEV). The second component represents joint variance that leads to variations of the hand path (non goal-equivalent variance, NGEV).; The first study involved drawing a circular pattern at a preferred frequency. In general, GEVNGEV was found suggesting the presence of error compensation. When motor abundance was limited, control of the hand's path became less stable due to high NGEV. In addition, control of the two hands as one linkage better reflected differences in the stability of bimanual patterns compared to control of the individual hand. The second study manipulated movement frequency of drawing and found a deterioration of the inter-limb synergy at higher movement frequencies when two arms were required to produce different patterns of muscle actions but not when two arms performed mirror-image actions. The third study demonstrated a potential link between the ability to account for inter-segmental dynamics and the organization of abundant joint motions underlying control of the hand path. For control of the dominant arm, more GEV and less NGEV were found when the effect of interaction torques was high. For control of the non-dominant arm, however, NGEV was found to be greater when such effect was high.