Research reportShort-term motor plasticity revealed in a visuomotor decision-making task
Introduction
Among the most challenging problems faced by our visuomotor system is the selection of targets in a cluttered world filled with many objects that could be acted upon. One possible solution to this problem is revealed by neurophysiological studies that suggest the brain plans multiple motor programs in parallel [1], [2], [3], [4], [5], allowing for several targets to compete for action selection at any one time. This strategy facilitates and simplifies target selection. Because all the potential actions are simultaneously coded, selection of the final action becomes the more straightforward process of one action plan ‘winning out’ over the others. This motor competition requires that each potential plan be associated with levels of activation reflecting its likelihood of being selected [e.g. [1]], but it is rather poorly understood what specific target and task properties modulate this competitive process.
In a recent experiment, we found that when subjects performed rapid reaches toward two equally likely targets before one target was cued in-flight, subjects initiated a ‘spatially’ averaged trajectory toward the midpoint of the potential target locations [6]. This is consistent with other eye-movement and reach paradigms demonstrating that movements made in the presence of competing stimuli tend to deviate between the stimulus locations [7], [8], [9], [10]. In a second experiment, we also showed that reach trajectories were biased both by the spatial location of the potential targets and by the number of targets on each side of space, suggesting that location and probability are factors that influence motor plan competition [6]. This pattern of results, however, is consistent with the idea that all the potential targets (and actions) have identical weights. Yet it is necessary to have a visuomotor system that can independently adjust the weightings of each potential target and action. Indeed, motor decisions are based not only on the information currently available to the sensory system, but also on previous visuomotor experience [e.g. [11]]. Visuomotor experience and the changes associated with that experience can accrue over timescales of weeks or months [12], [13] but can also be seen to operate over much shorter intervals [e.g. [9], [14], [15], [16]]. Specifically, it has been shown that the parameters of a current movement are influenced by the characteristics and intentions of the previous movement. It is precisely this trial-to-trial motor plasticity that we investigated in the current experiment: how the selection of a given target on one trial alters its weighting in the competition between targets on the subsequent trial.
Trial history effects reveal how visuomotor decision-making processes evolve across multiple movements [e.g. [14], [15], [17], [18]] and have been investigated in a wide range of paradigms. For example, subjects are faster to detect targets when the location is repeated [19], and grasping kinematics are affected by the recent availability of visual feedback [20], [21]. Target-directed eye movements provide a particularly good illustration of the effects of trial history: the colour, shape and location of the target on a previous trial influences both the neural activity in eye-movement-related structures during motor planning and the subsequent behavioural response [e.g. [7], [22], [23]]. In addition, behavioural studies have shown that reach targets embedded among distractors that vary in colour from trial-to-trial produce more variable reach trajectories than targets that maintain a colour across trials [24]. Of particular relevance is recent work investigating the effects of previously performed arm movements on subsequently performed actions. For example, avoiding a virtual obstacle on one trial will result in a more curved trajectory on the next trial even when no obstacle is present [25], [26]. These results suggest that successive movements in a sequence are not programmed de novo but instead are created by slightly modifying the blueprints of the preceding movement(s) [27]. In the current study, we wanted to capitalize on this proposed holdover of motor parameters from trial-to-trial as a way of creating a disparity in the weightings assigned to two potential targets. As such, we tested if the spatial averaging between potential targets that we have previously reported [6] would be biased toward the location of a previously cued target, and if so, whether this bias would accumulate across multiple trials.
Section snippets
Methods
We recorded rapid reach movements (OPTOTRAK, 150 Hz) from 17 right-handed (mean age 25.5 years, 10 female) subjects as they reached from a start button to a touch screen (40 cm away). Trials began with participants holding down the start button and fixating a cross centered on screen for a variable time ranging from 1000 to 2000 ms. A beep signalled when fixation was replaced by a target display, consisting of one or two outline targets (1-cm radius circle, black, on a white background), and also
Results
We employed functional data analysis techniques [28] to fit mathematical functions (using b-splines, see Supplemental Material and our previous work [6] for description of this technique) and to spatially normalize the reach trajectories. This enabled us to use functional analyses of variance (FANOVAs) to compare reaching behaviour across the conditions of interest. A FANOVA is a statistically sensitive technique which extends a traditional univariate ANOVA (which is conventionally used to
Discussion
Effective goal-directed behaviour in a dynamic environment requires a nervous system that is able to flexibly adapt its current behaviour based on prior experience. The ease with which the visuomotor system incorporates previous events into current motor plans demonstrates just how sophisticated and highly adaptable the underlying decision processes are [29]. Indeed, as we show here, earlier experience with particular targets can affect the weightings assigned to those and other targets
Acknowledgements
We would like to thank Dr. Paul Gribble for providing useful comments and suggestions as well as helping with the statistical analysis used in this study. This work was supported by operating grants from the Natural Sciences and Engineering Research Council to Jody Culham (Grant# 249877-2006 RGPIN) and Melvyn Goodale (Grant# 6313-2007 RGPIN).
References (36)
- et al.
Neural correlates of reaching decisions in dorsal premotor cortex: specification of multiple direction choices and final selection of action
Neuron
(2005) - et al.
Short-term priming, concurrent processing, and saccade curvature during a target selection task in the monkey
Vision Res
(2001) - et al.
Hidden cognitive states revealed in choice reaching tasks
Trends Cognit Sci
(2009) - et al.
Eye movement trajectories and what they tell us
Neurosci Biobehav Rev
(2006) - et al.
Bayesian decision theory in sensorimotor control
Trends Cognit Sci
(2006) - et al.
Trial-by-trial motor adaptation: a window into elemental neural computation
Prog Brain Res
(2007) Motor learning is optimally tuned to the properties of motor noise
Neuron
(2009)- et al.
A kinematic analysis of reaching and grasping movements in a patient recovering from optic ataxia
Neuropsychologia
(1991) - et al.
Target selection in visual search as revealed by movement trajectories
Vision Res
(2008) - et al.
Hand path priming in manual obstacle avoidance: rapid decay of dorsal stream information
Neuropsychologia
(2009)
The problem of serial order in behavior: Lashley's legacy
Hum Mov Sci
Attentional landscapes in reaching and grasping
Vision Res.
Dorsal premotor neurons encode the relative position of the hand, eye, and goal during reach planning
Neuron
Cortical mechanisms of action selection: the affordance competition hypothesis
Phil Transact R Soc Lond B
Representation of averaging saccades in the superior colliculus of the monkey
Exp Brain Res
Competition between saccade goals in the superior colliculus produces saccade curvature
J Neurophysiol
Modulation of neuronal activity by target uncertainty
Nature
Reaching for the unknown: Mutliple-target encoding and real-time decision-making in a rapid reach task
Cognition
Cited by (28)
Tracking priors and their replacement: Mental dynamics of decision making in the number-line task
2022, CognitionCitation Excerpt :We verified that this Condition effect was observed not only in the regressions, but also in the trajectories: at the time when the condition effect was at its peak (350 ms), the average implied endpoint was 41.0 for the left-bias condition, 47.4 for the unbiased condition, and 53.6 for the right-bias condition. Because the Condition predictor correlated with the target numbers (r = 0.32), its effect could have been arguably explained as perseveration from recent trials – e.g., because the motor system may be biased towards recent movement plans (Chapman et al., 2010b). However, the results clearly refute this alternative interpretation: the Condition effect could not be reduced to an effect of the recent target numbers, because it had a significant and strong effect although the regression model also included the targets of the 4 recent trials.
Attentional capture in goal-directed action during childhood, adolescence, and early adulthood
2022, Journal of Experimental Child PsychologyReaching for known unknowns: Rapid reach decisions accurately reflect the future state of dynamic probabilistic information
2021, CortexCitation Excerpt :The stimulus in the current experiments is separate from the target and conveys information about the probability of the final target location, rather than cueing location directly. In this task, we extend a previous go-before-you know paradigm (Chapman et al., 2010a, 2010b; Gallivan et al., 2017; Milne et al., 2013; Wood et al., 2011), which requires participants to initiate a movement in response to a go-signal before one of two potential target locations is revealed as the final target. Here, the probability of the upcoming target location was conveyed to participants via a stimulus that rotated at a fixed rate, either clockwise (CW), counter-clockwise (CCW), or, on baseline/control conditions, remained stationary (Fig. 1; see videos linked in Open practices).
Decision making in slow and rapid reaching: Sacrificing success to minimize effort
2020, CognitionCitation Excerpt :One might argue that our failure to observe clear averaging behaviour in Experiment 1 may be due to employing not only a much more liberal movement time criterion, but also a more liberal reaction time criterion (i.e. time available for movement planning). Previous studies by Chapman and colleagues that consistently found evidence for averaged movement trajectories constrained participants' reaction times to 325 ms after target presentation (e.g., Chapman et al., 2010a, 2010b). As our pilot studies indicated that most of our participants were unable to meet this reaction time criterion in most of the trials, we opted for a more generous reaction time limit of 500 ms. We think the reason that our participants needed more time for movement initiation can be attributed to the use of a slightly different setup.
Rapid Automatic Motor Encoding of Competing Reach Options
2017, Cell ReportsCitation Excerpt :On day 1 of testing, participants completed the pre-adaptation phase of the experiment, in which they performed one-target trials and two-target go-before-you-know trials. Previous studies have shown that, when initiating reaches toward two potential targets, many participants aim toward the midpoint and then make a correction toward the cued target (Chapman et al., 2010a, 2010b; Stewart et al., 2013, 2014; Gallivan and Chapman, 2014). However, one-third to one-half of all participants do not exhibit consistent trial-to-trial averaging behavior and, instead, will at least occasionally adopt a strategy that involves picking one of the two potential targets to aim toward (Stewart et al., 2013, 2014).
- 1
These authors contributed equally to this work.