Analyzing Perturbation Characteristics in an Asymmetrical Bimanual Movement

Many motor tasks in everyday life, such as driving and cooking, involve a combination of discrete and rhythmic movements. Although an increasing number of studies have identified discrete and rhythmic movements as dynamic primitives in complex motor control, specifying conditions in which rhythmic movements facilitate discrete movements and vice versa remains unknown. While previous studies have focused on bimanual interactions in movement preparation and initiation, the present study quantifies the magnitude and direction of perturbation during the execution of bimanual movement. Specifically, we measure the change of kinematics when ongoing rhythmic movement in one arm is perturbed by rhythmic or discrete initiation of the other arm. Fifteen healthy young adults participated in the study. The participants performed bimanual forearm rotation on the horizontal plane. In one condition (Discrete initiation condition), their non-dominant arm initiated a point-to-point reaching while their dominant arm moved rhythmically at a specific tempo (0.75 Hz). In the other condition (Rhythmic initiation condition), their non-dominant arm initiated a rhythmic movement while the dominant arm performed the same rhythmic movement as in the first condition. For both conditions, timing of the imperative visual cues for initiation was randomized. The elbow angle profile from the two forearm rotations was recorded by absolute encoders so that the veridical elbow angle was displayed on a monitor in real time. The task consists of two phases: Familiarization and test. In the familiarization phase, the participants practiced moving their forearm at a specific speed (180 deg/s) for the discrete movement and at a specific tempo (0.75 Hz) for the rhythmic movement unimanually and then bimanually. In the test phase, approximately 80 trials for the Discrete initiation condition (40 inward and 40 outward movements) and 80 trials for the Rhythmic initiation condition were performed for each participant. We analyzed the magnitude and direction of the perturbation in the ongoing rhythmic movement in the dominant arm as calculating the mean difference between the desired and actual profile of instantaneous phase in the dominant arm for the duration of a full cycle upon the initiation in the non-dominant arm. Results showed that, when a rhythmic movement is initiated, the amount of perturbation in the ongoing rhythmic arm depended on the phase relationship between the two arms at the initiation of the non-dominant arm. When a discrete movement is initiated, the perturbation in the ongoing rhythmic arm exhibited similar patterns, but the overall perturbation showed positivity regardless of the relative phase. Interestingly, the same analysis with a shorter duration (125 ms) revealed that discrete initiation did not show the characteristics that were shown in analyses for longer duration. It was also shown that the reaction time of rhythmic initiation was longer than that of discrete initiation. However, neither rhythmic nor discrete initiation condition showed phase dependency. Taken together, the results suggest that the underlying neural mechanisms for discrete movement partly share those for rhythmic movement. Nevertheless, the overall positivity observed in the perturbation pattern and failure to show phase dependency during first 125 ms upon initiation support the existence of a distinct neural underpinning for discrete movement.