Kinematic Markers of Distance-Specific Control in Linear Hand Movements

ABSTRACT

In this article, the authors analyze kinematic characteristics of reaching movements to memorized visual target locations. An increase in target distance was associated with a decrease in correlation between peak acceleration and movement distance and with a simultaneous increase in correlation between peak acceleration and movement time. According to the previous work on motor control in isometric force responses and in reaching movements these results seem to indicate a continuous transition from a rather preplanned to a more corrective mode of movement control, which may be associated with an adaptive mechanism serving to counteract an increase in signal-dependent noise of the motor system.

Keywords:

• arm movements
• motor control
• trajectory control

ACKNOWLEDGMENTS

This research was supported by grants from a German Research Foundation (DFG, Graduiertenkolleg 885—“Brain and Behavior”—and Forschergruppe 560—“Perception and Action”).

The authors thank an anonymous reviewer for helpful and constructive comments on an earlier version of this manuscript.

Notes

1. The minimum variance theory does not allow direct predictions in respect to online corrections of trajectory, as it was formulated to account for motor planning. However, it can explain Fitts's law as a consequence of signal-dependent noise (Harris & Wolpert, 1998). As suggested by Schmidt et al. (1979), Fitts's law may be assumed to well account for slower responses, in which error corrections are possible. In rapid movement tasks (with movement times shorter than 200 ms), an analogous speed accuracy tradeoff can be explained by the variability of the initial force pulse and without the assumption of feedback-based corrective processes. Accordingly, our hypothesis is not inconsistent with the minimum variance approach. Moreover, a similar assumption can be derived from the work of Meyer et al. (1988) incorporating Fitts's as well as Schmidt's laws.

2. The data presented was collected within a comprehensive study aiming at the investigation of kinematic and electrophysiological correlates of planning and control of linear hand movements. In a previous article we reported a part of the results of this study (Kirsch & Hennighausen, 2010). In particular, we described electrophysiological markers of planning and control of one-dimensional hand movements associated with a varying degree of motion and their relation to the averaged kinematic trajectories. For the present report we used data collected in this study and focused on a single trial analysis of selected kinematic parameters.

3. One of the aims of the overall study was to investigate a possible dependence of visuomotor mapping on time delay inserted between stimulus and response. Based on the work on the relation between vision for perception versus vision for action (e.g., Goodale & Milner, 1992), we were primarily interested in EEG indicators of early stages of visuomotor planning. The purpose of the present analyses was quite different. Here, we focused on low-level motor control mechanisms, which may be assumed to be widely independent from the response delay, but dependent on the movement magnitude. Nevertheless, we included the factor delay in all analyses in order to ensure that the results are valid for all delay conditions.

4. Figure 2A gives an impression that the relationship between targets and correlations between peak acceleration and movement time is positive. Note, however, in Figure 2 the respective y-axis is reversed so that more positive values are at the bottom. Thus, the observed negative slopes are fully consistent with the data of the main analysis.

5. A similar analysis was also performed by Gordon and Ghez (1987b) to ensure that compensatory adjustments improve response accuracy. They estimated peak force variability, if peak d²F/dt² were a perfect predictor of peak force. In their analysis, the coefficients of variability in peak force increased in respect to the observed values indicating that the overall accuracy was improved compared with a simulated case in which force rise time was constant.