Dissociating the structural and metrical specifications of bimanual movement

The Interaction of Tactile Information and Movement Amplitude in a Multijoint Bimanual Circle-Tracing Task: Phase Transitions and Loss of Stability

Adaptive behaviour in bimanual coordination was examined with the use of a bimanual circle-tracing task. Circle diameter and tactile information were manipulated to form four tracing conditions: tracing a pair of 3-cm diameter circles with the tips of the index fingers (3F) or hand-held styli (3S) and tracing a pair of 10-cm diameter circles with the tips of the index fingers (10F) or hand-held styli (10S). Movement frequency was increased in all conditions. In the 3F, 3S, and 10S tracing conditions, an abrupt transition from asymmetric to symmetric coordination was the main adaptive response, while in the 10F tracing condition, phase wandering was the main adaptive response. Enhancement of fluctuations in relative phase, a signature of loss of stability, occurred before the transition from asymmetric to symmetric coordination. Movement frequency and movement amplitude interact as control parameters in this task. The results are discussed with reference to tactile surface contact and joint motion as sources of sensory information that can be used to stabilize bimanual coordination patterns. The presence or absence of tactile information is directly linked to the specific form of adaptive behaviour (phase transition or phase wandering) that emerges as a function of required movement amplitude and required pacing frequency.

Structural Constraints on the Performance of Symmetrical Bimanual Movements with Different Amplitudes

In bimanual movements the amplitude of each hand's movement often depends on the concurrent amplitude of the other hand's movement such that both amplitudes become similar (amplitude coupling). We tested the hypothesis that the strength of amplitude coupling depends on the tempo of performance of a movement sequence, a hypothesis based on a model of bimanual coordination that holds that cross-talk occurs at the execution level as well as at the programming level. Subjects performed bimanual periodic arm movements on two digitizers. In nine conditions constant small, constant large, and alternating small and large amplitudes of each arm were orthogonally combined. Overall tempo was varied by instructing subjects to increase the tempo progressively by 10%. Clear tempo-dependent modulations of the amplitude were observed in movements with instructed constant amplitude when the other hand performed alternating amplitudes. The effect of the size of constant-amplitude movements on the mean amplitude of the other hand indicated cross-talk at the execution level. Cross-talk at the programming level was revealed by the dependence of the current amplitude on the change in the amplitude of the other hand in the preceding cycle. Finally, asymmetric cross-manual effects were observed.

Shared bimanual tasks elicit bimanual reflexes during movement

Previous research has suggested distinct predictive and reactive control mechanisms for bimanual movements compared with unimanual motion. Recent studies have extended these findings by demonstrating that movement corrections during bimanual movements might differ depending on whether or not the task is shared between the arms. We hypothesized that corrective responses during shared bimanual tasks recruit bilateral rapid feedback mechanisms such as reflexes. We tested this hypothesis by perturbing one arm as subjects performed uni- and bimanual movements. Movements were made in a virtual-reality environment in which hand position was displayed as a cursor on a screen. During bimanual motion, we provided cursor feedback either independently for each arm (independent-cursor) or such that one cursor was placed at the average location between the arms (shared-cursor). On random trials, we applied a 40 N force pulse to the right arm 100 ms after movement onset. Our results show that while reflex responses were rapidly elicited in the perturbed arm, electromyographic activity remained close to baseline levels in the unperturbed arm during the independent-cursor trials. In contrast, when the cursor was shared between the arms, reflex responses were reduced in the perturbed arm and were rapidly elicited in the unperturbed arm. Our results thus suggest that when both arms contribute to achieving the task goal, reflex responses are bilaterally elicited in response to unilateral perturbations. These results agree with and extend recent suggestions that bimanual feedback control might be modified depending on task context.

Discrete bimanual co-ordination in children and young adolescents with hemiparetic cerebral palsy: Recent findings, implications and future research directions

Bimanual co-ordination is a field that has generated much research interest. It is clear that when hands move simultaneously there is a tendency for such movements to be synchronized. This phenomenon has also been observed in individuals with congenital brain damage, such as cerebral palsy. This paper provides an overview of past and current work in the area of bimanual co-ordination looking at the nature of the coupling and the underlying control of such movements. Work on 'non-disabled' individuals and unimanual reaching and grasping is included, but the main emphasis is bimanual co-ordination in children with hemiparetic cerebral palsy. The direction for future research is discussed and consideration given to the therapeutic implications of such work.

The influence of perturbing the working surface during reaching and grasping in children with hemiplegic cerebral palsy

PURPOSE

To examine unimanual and bimanual reaching and grasping in children with hemiplegic cerebral palsy with particular emphasis on the nature and extent of interlimb coupling when the working surface is perturbed.

METHOD

Nine children with hemiplegic cerebral palsy and 7 control children with no movement difficulties took part in the study. Children were asked to pick up a cube unimamually and bimanually when the surface it was placed on was either sloping away from the child (Experiment 1) or towards the child (Experiment 2). Both 3D kinematic data and video data were gathered and qualitative descriptions of video data were made.

RESULTS

The working surface did indeed influence the nature and extent of interlimb coupling and this varied from participant to participant. Analysis of the displacement data revealed that during the bimanual condition lower trajectories were produced by both the hemiplegic and non hemiplegic sides, especially in Experiment 2. The control group showed little difference between the unimanual and bimanual condition.

CONCLUSIONS

Evidence of interlimb coupling is found, these studies support the findings of our previous work that indicates that there are some benefits to performing bimanual movements in children with hemiplegic cerebral palsy.

Bimanual training reduces spatial interference

The authors investigated whether training can reduce bimanual directional interference by using a star-line drawing paradigm. Participants (N = 30) were required to perform rhythmical arm movements with identical temporal but differing directional demands. Moreover, the effectiveness of part-task training in which each movement was practiced in isolation was compared with that of whole-task training in which only combined movements were performed. Findings revealed that bimanual training substantially reduced spatial interference, but unimanual training did not. The authors therefore concluded that the spatial coupling of the limbs is not implemented in a rigid way; instead, the underlying neural correlate can undergo plastic changes induced by training. Moreover, the practical implication that emerged from the present study is that athletic, musical, or ergonomic skills that require a high degree of interlimb coordination are best served by whole-task practice.

Kinematics of a coordinated goal-directed bimanual task

The experiments address the problem of bimanual coordination in a familiar task of everyday life. A goal-directed drawer-pulling task, with asymmetrical assignments among hands, was analyzed with the objective to detect discrete kinematic events ('anchors') that potentially could serve in proper goal synchronization. The left hand reached out for the drawer and opened it while the right hand performed a prehension movement to pick up a peg from the drawer. The task was smoothly performed, independently of vision. Typically, trajectories and velocity profiles of the leading pull-hand were more stereotypical than the more variable ones of the pick-hand. The pull-hand had a large velocity peak during reaching, followed by a small peak during pulling. Velocity profiles of the pick-hand were not bell-shaped and exhibited one or two broad waves, often with an irregular and probing evolution. Velocity profiles of both hands were aligned with the first or the second velocity peak of the leading pull-hand. In the majority of cases, temporal associations of events in the kinematics of the two limbs could thus be identified, which could serve to synchronize the hands at the goal. The nearly straight biphasic reach-and-pull trajectory of the leading hand contrasted with the more curved trajectory of the right pick-hand whereas, in the same unimanual action, the latter trajectories were quasi-rectilinear. Changing constraints (no vision, cutaneous anesthesia of pulling fingers) could change the coordination pattern. We argue that bimanual coordination relies on two interacting mechanisms: (1) feedforward control on the basis of sensorimotor memory; (2) temporal adjustments during the evolving bimanual synergy. Multiple strategies, imposed by the leading pull-hand, appeared to be responsible for feedback-induced corrections in the pick-hand and were found to contribute to the goal-invariance and to the principle of motor equivalence.

Bilateral facilitation of motor control in chronic hemiplegia

The present study addressed the efficacy of concurrently moving both arms, with and without a load added to the uninvolved arm, in facilitating the quality of movement of the involved side in individuals with moderate, chronic hemiplegia. Six hemiplegic cerebrovascular accident (CVA) subjects with left-hemisphere lesions participated in the study. The four males and two females ranged from 46 to 77 years of age and 30-96 months post-CVA. All subjects scored at least 70% on the Fugl-Meyer test of motor function. The task was to perform discrete unilateral and bilateral elbow extensions in the horizontal plane. The movements were 45 degrees in amplitude and were to terminate in a 10 degrees target zone that was indicated by an illustration of a coffee mug. The instructions were to move toward the mug as smoothly as possible in a movement time (MT) determined to be 20% longer than their minimal MT for that distance. The primary dependent variable was the percentage of continuous vs. discontinuous trajectories observed in each condition, based on whether or not a transient hesitation or reversal was observed. Phase of peak velocity was also quantified as a general indication of the symmetry of the velocity profile. Three of the six subjects exhibited a greater percentage of continuous movements of the involved arm in the nonloaded bilateral condition than the unimanual condition. Five subjects benefited when the uninvolved arm was inertially loaded in the bilateral condition when compared with unimanual performance. Only the oldest subject failed to exhibit facilitation. Peak velocity phase tended to normalize toward symmetry in the bilateral conditions. These findings are consistent with prior evidence that the control of the involved arm improves during bimanual performance for some hemiplegic subjects. It further suggests loading the uninvolved arm may benefit some subjects with respect to unimanual performance, with age perhaps playing a role in determining efficacy.

Generation of bimanual trajectories of disparate eccentricity: levels of interference and spontaneous changes over practice

The authors examined intermanual interactions of 2 hands that were required to concurrently follow trajectories that differed in eccentricity. Ten healthy participants attempted to learn to trace 2 figures, a circle and an ellipse, with bilaterally isochronous (1:1) timing demands. Initial unimanual trials were followed by bilateral practice comprising 750 movement cycles. Two objectives were addressed: The authors' primary aim was to determine if kinematic interlimb interference is evident independent of spatial and temporal interference and to observe the potential practice-related changes in the nature of that interference. That test was afforded by participants' natural tendency to draw a circle with a relatively constant tangential velocity and an ellipse with a systematically varying velocity. A second aim was to observe the nature of spontaneous changes in the performance of each individual effector, and in the relationship between effectors, across practice. Those objectives were specifically addressed in a context in which augmented feedback was not available to direct the learners' attention to a particular feature of performance. The results suggested that interlimb assimilation of spatial features is the primary source of interference for that task and that apparent effects at the kinematic level are the secondary, indirect product of spatial coupling. Those results were found across blocks of practice. With respect to nondirected performance changes, substantially less improvement was evident in the performance of each individual effector than in the reduction of interlimb interference. Specifically, no practice-related changes in temporal variability or velocity bias, and minimal changes in trajectory smoothness, were evident in individual limbs. Conversely, significant reductions were observed in the variability of relative phase between limbs and in the magnitude of interlimb phase lag.

Bilateral reaching to asymmetrical targets: muscle and joint dynamic interlimb adaptations

A combined analysis of time, electromyographic, and joint torque measures was used to explore the force control processes underlying the dissociation of arm reaching movements performed bilaterally to targets of varying amplitude. Limb movements appeared closely coupled at movement initiation, which was confirmed by a strong tendency of the agonist muscles to remain synchronized despite any interlimb asymmetry in final target distance. On the other hand, interlimb decoupling occurred later as a result of the difference in the antagonists' timing of activation between the limbs. The partitioning of the net joint torque revealed that muscle activity is regulated in response to the intersegmental dynamics of the limb. It is proposed that spatial decoupling of asymmetrical movements becomes possible through postinitiation feedback processes which regulate muscle recruitment phenomena.

Coping with systematic bias during bilateral movement

The present studies examined the nature of kinematic interlimb interference during bilateral elbow movements of 1:1, 2:1 and 3:1 frequency ratios and the manner in which subjects cope with coordination bias. Analysis of movement trajectories in the first experiment indicated progressively greater angular velocity assimilation across 2:1 and 3:1 conditions. The desired temporal relationship was maintained by slowing or pausing the low-frequency movement at peak extension while the high-frequency arm produced intervening cycles. An increase in amplitude was also evident for concurrent, homologous cycles. Movement smoothness was emphasized and additional practice was provided in a second experiment. This resulted in dissociated peak angular velocity between limbs and eliminated hesitations and amplitude effects. Bias was still evident, however, as an intermittent approach toward a 1:1 ratio within each cycle. This systematic tendency was somewhat greater at the lower of two absolute frequency combinations but was not influenced by the role of each arm in producing the higher or lower frequency movement. The findings from the first experiment suggest that subjects initially accommodate interlimb kinematic assimilation, while producing the intended timing ratio, by intermittently slowing or pausing the lower-frequency movement. This attenuates the need for bilaterally-disparate movement parameters and provides additional time for organizing residual kinematic differences, perhaps reducing "transient coupling." Evidence from the second experiment indicates that subtle relative motion preferences are still evident following sufficient practice to perform the movements smoothly. The within-cycle locations of the points of greatest interlimb bias for the 2:1 rhythms were positively displaced from those previously observed for 1:1 oscillations. The persistent coordination tendencies noted in both experiments perhaps reflect an assimilation/compensation cycle and constitute one potential source of the systematic error that often emerges during the acquisition of complex skills.

Catching With Both Hands: An Evaluation of Neural Cross-Talk and Coordinative Structure Models of Bimanual Coordination

Kelso, Southard, and Goodman (1979) and Marteniuk and MacKenzie (1980) have each proposed a different theoretical model for bimanual coordination. In the model of Kelso et al., a close temporal relationship between the hands in a bimanual task is predicted, even when each hand is required to move different distances. In Marteniuk and MacKenzie's model, separate motor commands are issued so that each limb will arrive simultaneously at the specified movement endpoint, leading to low temporal associations between limbs. In most previous work on bimanual coordination, manual aiming tasks with differing constraints have been used by subjects in individual studies. In this study, the usefulness of existing models for predicting performance in a real-world catching task in which the required movement pattern was constrained by ball flight characteristics was examined. Eleven university students caught tennis balls with both hands in the following 3 conditions: Condition 1. Ball projected to the right shoulder area (left hand moved a greater distance than the right); Condition 2. Ball projected to center of the chest area, (both hands moved same distance); and Condition 3. Ball projected to left shoulder area (right hand moved a greater distance). Kinematic data (time to peak velocity, movement initiation time) indicating significant cross-correlations between the left and right limbs in all 3 conditions concurred with the data of Kelso et al. (1979) on manual aiming. Timing appeared to be an essential variable coordinating bimanual interceptive actions. Although the limbs moved at different speeds when each was required to move different distances, times to peak velocity showed strong associations, suggesting the presence of a coordinative structure.

Spatial coupling in the coordination of complex actions

The majority of investigations on coordinated action have focused on temporal constraints in movements. Recent studies have demonstrated spatial constraints when the hands produce different trajectory shapes simultaneously. The focus of the current study was to determine whether spatial coupling occurs in individual parameters of the actions, or whether the shapes per se undergo accommodation. Subjects were tested on a bimanual paradigm to investigate the nature of spatial constraints in complex tasks. Shape and size of the required trajectories were varied for two limbs. When trajectories that require different shapes were assigned to the two hands, disruption in the spatial characteristics of the trajectories was observed. Disruption in the global patterns of the trajectories could be described on the basis of coupling in individual parameters of action, direction, and amplitude, which could be inferred by decomposing the trajectories into orthogonal components. Amplitude accommodation in these orthogonal components of motion increased linearly with the difference in required amplitude for the two limbs. Interpretations of these effects suggest that directional coupling is a result of interference between two different response plans, whereas amplitude coupling may be related to either planning or execution variables. These results strongly suggest the need for further investigation of the spatial domain of complex coordinated action.

Between-limb asynchronies during bimanual coordination: effects of manual dominance and attentional cueing

Whereas previous studies on interlimb coordination have mainly underscored the ubiquitous tendency to synchronize the motions of the limbs, the present experiment revealed a small, but distinct, interlimb asynchrony or phase offset, i.e. the dominant limb led the non-dominant limb during the production of bimanual circle drawing. This asynchrony was clearly evident in the majority of right-handers, but not in left-handers. Moreover, attentional cueing affected the size of the asynchrony. Instructions to visually monitor the dominant limb or non-dominant limb strengthened and weakened the phase offset, respectively. A multifactorial neural account is proposed to underly the temporal asynchrony.

Interlimb coupling in children with hemiplegic cerebral palsy

Children with hemiplegic cerebral palsy participated in three experiments designed to examine interlimb coupling during reaching and pointing/grasping tasks. Comparisons were made between the hands working in isolation and working together. Descriptions of the movements involved both qualitative and quantitative methods, using normal video analysis and kinematic analysis. In order to obtain a total picture of reaching and grasping it was necessary to examine multiple variables both qualitatively and quantitatively. There was strong evidence of coupling between the hands, with coupling of timing being the most prevalent followed by coupling of trajectory and hand posture. All children showed some form of coupling in the experiments.

Relative phase destabilization during interlimb coordination: the disruptive role of kinesthetic afferences induced by passive movement

The disruption of three patterns of two-limb coordination, involving cyclical flexion-extension movements performed in the same or in different directions, was investigated through application of passive movement to a third limb by the experimenter. The three patterns referred to the homologous, homolateral, and heterolateral (diagonal) limb combinations which were performed in the sagittal plane. The passive movement involved a spatiotemporal trajectory that differed from the movements controlled actively. Even though subjects were instructed to completely ignore the passive limb movement, the findings of experiment 1 demonstrated a moderate to severe destabilization of the two-limb patterns, as revealed by analyses of power spectra, relative phase, cycle duration, and amplitude. This disruption was more pronounced in the homolateral and heterolateral than in the homologous effector combinations, suggesting stronger coupling between homologous than nonhomologous limb pairs. Moreover, passive mobilization affected antiphase (nonisodirectional) movements more than inphase (isodirectional) movements, pointing to the differential stability of these patterns. Experiment 2 focused on homolateral coordination and demonstrated that withdrawal of visual information did not alter the effects induced by passive movement. It was therefore hypothesized that the generation of extra kinesthetic afferences through passive limb motion was primarily responsible for the detriment in interlimb coordination, possibly conflicting with the sensory information accompanying active movement production. In addition, it was demonstrated that the active limbs were more affected by their homologous passive counterpart than by their nonhomologous counterpart, favoring the notion of "specific" interference. The findings are discussed in view of the potential role of kinesthetic afferences in human interlimb coordination, more specifically the preservance of relative phasing through a kinesthetic feedback loop.