Exp Brain Res. 2015 Jul;233(7):2141-53. doi: 10.1007/s00221-015-4285-x.
Influence of workspace constraints on directional preferences of 3D arm movements.
- Kinesiology Program, Arizona State University, 500 N. 3rd Street, Phoenix, AZ, 85004, USA.
We previously demonstrated a tendency to perform arm movements by using a trailing joint control pattern during which either the shoulder or elbow is rotated actively and the other (trailing) joint is rotated predominantly passively, by interaction torque during horizontal movements and by interaction and gravitational torque during 3D arm movements. This tendency was established with a free-stroke drawing task that required production of series of strokes in randomly selected directions from the center to the perimeter of a horizontal circle. The studies demonstrated that within a planar workspace, the usage of the trailing pattern depends on movement direction and the most frequently selected directions are those providing the opportunity to use the trailing pattern. Here, we studied whether the opportunity to use the preferred trailing pattern also depends on the orientation of the planar workspace. The free-stroke drawing task was performed with unconstrained arm movements within circles of a left-diagonal (LD) and right-diagonal (RD) orientation. Two pronounced preferred directions were revealed in the LD condition, and they were the directions in which the trailing pattern was used. Directional preferences were less pronounced, and the trailing pattern was not observed in any directions in the RD condition. Also, subjects identified the RD condition as inconvenient. The results reinforce the previous finding of the propensity to use the trailing pattern during arm movements. They also suggest that orientations of the workspaces in manual activities should be designed to support the trailing pattern as a favored type of joint control.
Human limbs have complex biomechanical properties. Our research is driven by a hypothesis that the brain exploits these properties to make our movements more efficient. Gravity is an obvious mechanical factor influencing our movements. There are also less apparent but very influential factors caused by the multi-joint structure of our limbs. The limbs are linkages of consecutively connected segments. For example, the arm includes the upper arm connected to the trunk at the shoulder joint, the forearm connected to the upper arm at the elbow joint, and the hand connected to the forearm at the wrist joint. When we activate muscles spanning one of the joints, rotation of the segments at this joint causes the other segments to move too. For example, activation of shoulder muscles results in rotation of the upper arm relative to the trunk. However, the elbow and wrist would also rotate because the upper arm pulls the distal segments. The only way to prevent the rotation of the elbow and wrist would be activation of muscles at these joints to resist the mechanical effect of shoulder rotation.
In motor control research, the cause of rotation of the elbow and wrist in our example is termed “interaction torque” because this torque emerges due to mechanical interactions of the limb segments. A simple way to experience interaction torque is to rotate the elbow and keep the wrist muscles relaxed, as shown in Fig. 1. Although wrist muscles are not activated, the wrist rotates due to interaction torque caused by elbow motion. The wrist can be stabilized and the hand can be kept aligned with the forearm only through activation of wrist muscles that would generate muscle torque that copes with interaction torque.
Interaction torque is “passive” because at each joint, it is not caused by activity of the muscles spanning that particular joint. Interaction torque is complex because it depends on motion of all limb joints, and this dependence is not linear. Interaction torque is relatively high because it is proportional to accelerations and squires of velocities of other joint rotations. And yet, the brain needs to generate control signals to the muscles that would result in muscle torque that complements interaction and other passive torques at each joint and results in goal-directed movements of the entire limb. Complexity of this control problem is evident from the complexity of interaction torque and the large number of joints that are involved in natural movements of daily activities.
Our research suggests that the brain solves the complicated problem of control of multi-joint limbs in a simple way. Instead of coping with and regulating interaction torque at all joints, the nervous system uses a “trailing joint” control strategy consisting in generation of interaction torque that produces the required movement. Specifically, a single (we term it “leading”) joint is used to generate interaction torque that rotates the rest of the joints (“trailing” joints) in a way required by the task – akin to the handle that brings in motion the entire whip. This control strategy is described in detail in our reviews (Dounskaia 2005; 2010; Dounskaia and Shimansky 2016). Research also shows that although trailing joint muscles can intervene and modify motions of these joints caused by passive torques, the preference is to minimize this intervention. Accordingly, skillful movements of athletes and musicians are characterized by less interference with passive torques from active control at the trailing joints compared with novices.
One experimental approach that demonstrated the preference to allow the trailing joints to be moved predominantly passively was a free-stroke drawing task, which was also used in the present study. Experiment participants were presented a flat service with a circle depicted on it and asked to produce series of strokes with their fingertip (leaving no traces) from the center to the perimeter of the circle. They were asked to select directions of consecutive strokes randomly and at the same time to produce strokes in as many different directions as possible. The later instruction encouraged a uniform distribution of strokes around the circle. We performed this experiment using different circle orientations. In the present study, two slanted orientations were used, left- and right-diagonal. Horizontal orientations at the shoulder and waist levels were examined in our previous studies (Goble et al. 2007; Dounskaia and Wang 2014).
Despite the encouragement to distribute strokes evenly across the circle, participants demonstrated pronounced directional preferences by producing strokes more frequently in some directions and avoiding some other directions. The preferred directions were consistent across participants, although they were specific for each circle orientation and position. Profound experimental and analytical examination demonstrated that the only factor that could account for the directional preferences in all conditions was the tendency to minimize muscle torque at the trailing joint. In other words, the preferred directions were those which could be reached either by actively rotating the shoulder and allowing the elbow to rotate mainly passively or by actively rotating the elbow and allowing predominantly passive motion at the shoulder. Participants voluntarily kept the wrist fixed in all conditions.
What is the advantage of the trailing joint control pattern that makes it preferred? Since the trailing joints are rotated largely passively, one can think that low muscle effort is the advantage of the trailing pattern. Our research suggests that the tendency to reduce muscle effort is not the reason, or at least not the major reason why this pattern is preferred. The major reason is that this pattern represents a simplified neural control strategy (Dounskaia and Shimansky 2016). The simplicity of joint control and coordination provided by this pattern releases more neural resources for performance of concurrent cognitive tasks. For example, the use of passive torques for joint rotation allows a pianist to concentrate attention on tiny adjustments of the stroke that improve the produced sound and a tennis player to focus on the ball and movements of the partner, which is necessary for a good shot.
The finding of the preferences to use the trailing joint control pattern to perform limb movements and understanding of its advantages has immense practical importance in many areas, from health and ergonomics to sports science. For example, this finding should be taken into account during development of working environment. If the activity requires frequent reaching for certain objects, these objects should be positioned to allow the use of the trailing pattern during reaching. This would improve accuracy of performance and reduce both muscular and mental fatigue. The present study demonstrates that spatial constraints imposed by the environment can limit the opportunities for the use of the trailing joint control pattern. Thus, the environment should be examined and such constraints should be minimized. Prevention of movements that require substantial muscular intervention at the trailing joints would also optimize man-computer interactions, from common mouse to touchscreens and emerging touchless gesture interfaces. In addition to accuracy, resistance to fatigue, and decreased risk for musculoskeletal disorders, prevention of non-preferred limb movements would support the use of the technologies by vulnerable populations, such as older adults, by reducing psychomotor, attention, and adaptability challenges. The importance of investigating whether required movements can be performed through the trailing pattern is emphasized by a tendency to simplify visual information by operating with simple, regular shapes which often require non-preferred movements in terms of joint control. Examples are drawing a circle, writing printing rather than cursive letters, and even moving the hand in the lateral direction. Thus, non-preferred movements are not always apparent, and a specialized examination of movements involved in each activity type is necessary.
The revealing of the trailing joint control strategy also creates new means for better understanding of neurological disorders, such as Parkinson’s disease, stroke, developmental disorders, etc. Albeit this control strategy is simplified, it emphasizes reliance of movement control on sophisticated neural processes. Similar to extensive practice needed to learn to crack a whip skillfully and precisely, this strategy more so emerges from intense motor learning and requires maintaining this knowledge and fine-tuned usage of it. Each neural disorder can affect this control process in a specific way, revealing of which would be insightful about the disorder.
Finally, it is expected that the trailing joint control strategy is not limited to arm movements but is also used during complex whole-body movements like many sports actions. Understanding of which part of the body is used as leading, the mechanical effect the leading body part produces on the rest of the body, and how musculature at the other body parts fine-tunes their movements would tremendously simplify learning of many complex sports activities. Although little research has been conducted in this direction, this expectation is supported by personal experience of the author Natalia Dounskaia whose alpine skiing skill skyrocketed and became much more effortless and safe after she deciphered the leading joint and the mechanical effect it has to generate on the rest of the body during down-hill skiing turns.
The provided examples show that the information obtained in the present study is an important step toward deep understanding of the preferred strategy of limb movement control that potentially can help improve effectiveness and safety of many aspects of human motor activities in various settings.
Figure 1. Cyclic flexion/extension of the elbow while wrist muscles are relaxed. Interaction torque generated by elbow rotation causes the wrist to flex and extend.
Dounskaia (2010) Control of human limb movements: the leading joint hypothesis and its practical applications. Exerc Sport Sci Rev, 38:201-208.
Dounskaia N (2005) The internal model and the leading joint hypothesis: Implications for control of multi-joint movements. Exp Brain Res, 166:1-16.
Dounskaia N, Shimansky Y (2016) Strategy of arm movement control is determined by minimization of neural effort for joint coordination. Exp Brain Res. 234:1335-1350.
Dounskaia N, Wang W (2014) A Preferred Pattern of Joint Coordination during Arm Movements with Redundant Degrees of Freedom. J Neurophysiol. 112:1040-1053.
Goble JA, Zhang Y, Shimansky Y, Sharma S, Dounskaia NV (2007) Directional biases reveal utilization of arm’s biomechanical properties for optimization of motor behavior. J Neurophysiol 98:1240–1252.