Work

Adaptive Control of Goal-oriented Human Arm Movement

Year: 2003
Type: article
Abstract: Experiment 1 directly tested a fundamental prediction of the equilibrium point hypothesis (EPH) of movement control, namely, equifinality.Subjects learned wrist flexion movements while a motor provided assistance in proportion to the instantaneous velocity.Subjects stopped short of the target on the trials where the magnitude of the assistance was randomly decreased, i.e., equifinality was not achieved.This is contrary to the EPH, although such effects are entirely consistent with predictions based on the formation of an internal dynamics model.In experiment 2 subjects performed 0.25 m horizontal reaches while exposed to a position-dependent parabolic field (PF), located in a region extending to 0.1 m from the start position.Following extensive learning of the PF, the field was randomly perturbed to 213 or 312 times the original strength for 5 trials, a total of 24 times.Mechanical channel trials were used to test for evidence of internal model formation.Internal model formation was evident as early as the second or third PF trial.Subjects produced lateral force to counteract the field, although it did not perfectly match the PF force, resulting in bowed trajectories.As learning progressed, we found significant decreases in the EMG in all 6 shoulder and elbow muscles once subjects had passed through the force field.We suggest subjects modified their internal model to more accurately compensate for the PF, such that the cocontraction could be reduced.Subjects quickly readjusted to the PF following each APF set, suggesting retention of the motor commands required in the PF.On the first APF trial of each set, subjects were displaced in the direction of the change in force.On the second trial, trajectories were substantially straighter, but the displacement was not reduced significantly more in the 3'd-5'h trials.Subjects adapted the internal model to more accurately compensate for higher strength APF, while simultaneously employing higher levels of cocontraction.In the lower strength APF results suggest a reduction in the level of cocontraction.We found no consolidation of learning between APF sets, i.e., performance in the APF did not improve as subjects were exposed to more sets.We hypothesized that when subjects adapted to the novel field the IM would be become more accurate at compensating for the applied field, with an associated drop in the level of cocontraction.We aimed to find out how quickly the IM begins to develop when subjects encounter the perturbed field.The objectives of this work are therefore twofold.Firstly, we aimed to elucidate whether EP or IM control is utilised by the CNS.Single joint movements are sufficiently simple that predictions can be made to distinguish the contrasting control theories.Secondly, a multi-joint task was used to present the CNS with a more complex control problem than a single joint task, thereby providing greater insight into the general principles of adaptation and motor learning.It is hoped that the two experiments within this thesis will lead to a clearer understanding of which control mechanisms the CNS uses in goaloriented movements, and how these develop as learning occurs in novel environments. Control by Internal Model or EPH?Studies of human muscular control often use learning paradigms in which subjects are asked to make task-oriented movements.Subjects learn to move to a target in a null field, where no external force is applied to the system.A novel force field, to which subjects have not previously been exposed, is then randomly, or consistently, introduced to perturb the learnt hand path trajectory.Analysis of adaptation to these forces over time, and the aftereffects that occur when the field is subsequently removed, can be measured using both kinematic and EMG measures.Patterns in learning can be used to deduce how the CNS modifies motor commands to improve performance.The results of such studies suggest three possible schemes, although not necessarily mutually exclusive, which may be employed by the CNS in controlling these goal-directed tasks: Internal Model formation, EP Control or Impedance Control.Present research continues in this area of neuromuscular control in an attempt to unambiguously determine whether or not the aforementioned mechanisms are used in particular tasks.The experiments within this thesis aim to more clearly elucidate the mechanisms that are used be the CNS in adapting to novel dynamic environments.
Language: English
Open Access status: green
OpenAlex ID: W2149688077
FWCI:
Cites: 47
Cited by:
Related to: 9