Review
The computational and neural basis of voluntary motor control and planning

https://doi.org/10.1016/j.tics.2012.09.008Get rights and content

Optimal feedback control (OFC) provides a powerful tool to interpret voluntary motor control, highlighting the importance of sensory feedback in the control and planning of movement. Recent studies in the context of OFC have increasingly used mechanical perturbations and visual shifts to probe voluntary control processes. These studies reveal the surprising sophistication of corrective responses, which are goal-directed and exhibit knowledge of the physical properties of the limb and the environment. These complex feedback processes appear to be generated through transcortical feedback pathways. The research reviewed here opens and enhances several lines of discovery, including testing whether feedback corrections share all of the attributes associated with voluntary control, identifying how prediction influences optimal state estimation, and importantly, how these voluntary control processes are generated by the highly distributed circuitry within the brain.

Section snippets

Changing views on the use of sensory feedback for voluntary control

It is easy to recognize that sensory signals, such as from our eyes, skin or muscles, help us perceive our body and the world around us. It is harder to understand how these same sensory signals help guide and correct our body movements. Sensory feedback is clearly involved when someone accidentally bumps your arm at a cocktail party leading to rapid motor corrections to avoid spilling a drink held in the hand. Less obvious is whether and how sensory feedback is used continuously to adjust

Optimal Feedback Control (OFC) as a theory of voluntary control

Optimality principles have been common in motor control with various possible objectives, such as minimizing jerk, torque change, or the influence of noise [4]. In 2002, Todorov and Jordan [5] proposed OFC as a theory of voluntary control (Figure 1a). Delayed sensory feedback that makes servo-control unstable is overcome by using optimal state estimation, a Kalman filter integrating efference copy signals with delayed sensory feedback. The derivation of the optimal control policy (i.e.,

Perturbations as a probe of voluntary control

Because feedback is an essential feature of OFC, a key approach to probe control is through perturbations, either visual or mechanical. The use of mechanical perturbations is not new – this approach has been used for many decades to observe muscle stretch responses (Box 1). What OFC clarifies is that control is highly specific to the ongoing behavioural goal. Perturbations do not just elicit stereotypical motor reflexes, but also voluntary control processes directly [21]. Thus, perturbations

Spinal and transcortical feedback

As a theory to describe behaviour, OFC is agnostic as to how voluntary motor control is generated by the spinal cord and brain. The spinal cord provides the first level of feedback processing. In many species, spinal circuits support sophisticated control, such as scratching, wiping, and basic locomotor patterns 36, 61, 62. Analyses of these movements highlight many of the key characteristics predicted by OFC, notably success with variability. The spinal cord also provides phase-dependent

Mapping OFC onto brain circuits

OFC-like control is speculated as being generated by cortical and sub-cortical brain circuits 21, 80. Figure 1b provides an overview of the putative contributions of various brain regions to voluntary control. It is at best a sketch that will evolve as experimentation helps to delineate further the contribution of different brain regions to control. The three basic processes, task selection, control policy, and state estimation, are each generated by highly distributed cortical circuits. The

Concluding remarks

The use of OFC to interpret voluntary movement has emphasized the importance of sensory feedback for control. Visual and mechanical perturbations during motor behaviour constitute an important scientific approach to probe the sophistication of feedback processing, providing a window into voluntary control processes. Recent studies reveal that corrective responses are highly adaptable based on the behavioural goal and consider the many complexities inherent in multi-joint movements. The basic

Acknowledgements

This work was supported by grants from the National Science and Engineering Research Council of Canada, Canadian Institutes of Health Research (CIHR), and a GlaxoSmithKlein-CIHR chair in Neurosciences. The author would like to thank members of the Limb Lab for helpful feedback on this article.

References (90)

  • E. Todorov et al.

    Optimal feedback control as a theory of motor coordination

    Nat. Neurosci.

    (2002)
  • E. Todorov

    Optimal control theory

  • A.A. Faisal

    Noise in the nervous system

    Nat. Rev. Neurosci.

    (2008)
  • W. Li et al.

    Iterative linearization methods for approximately optimal control and estimation of non-linear stochastic systems

    Int. J. Control

    (2007)
  • D.W. Franklin et al.

    Specificity of reflex adaptation for task-relevant variability

    J. Neurosci.

    (2008)
  • A.J. Nagengast

    Optimal control predicts human performance on objects with internal degrees of freedom

    PLoS Comput. Biol.

    (2009)
  • F.J. Valero-Cuevas

    Structured variability of muscle activations supports the minimal intervention principle of motor control

    J. Neurophysiol.

    (2009)
  • I.H. Stevenson

    Bayesian integration and non-linear feedback control in a full-body motor task

    PLoS Comput. Biol.

    (2009)
  • J.B. Dingwell

    Do humans optimally exploit redundancy to control step variability in walking?

    PLoS Comput. Biol.

    (2010)
  • A.G. Feldman et al.

    The equilibrium-point hypothesis--past, present and future

    Adv. Exp. Med. Biol.

    (2009)
  • G.E. Loeb

    Optimal isn’t good enough

    Biol. Cybern.

    (2012)
  • T.E. Hudson

    Compensation for changing motor uncertainty

    PLoS Comput. Biol.

    (2010)
  • D.A. Kistemaker

    The central nervous system does not minimize energy cost in arm movements

    J. Neurophysiol.

    (2010)
  • A. De Rugy

    Muscle coordination is habitual rather than optimal

    J. Neurosci.

    (2012)
  • V. Mohan et al.

    Passive motion paradigm: an alternative to optimal control

    Front. Neurobot.

    (2011)
  • S.H. Scott

    Optimal feedback control and the neural basis of volitional motor control

    Nat. Rev. Neurosci.

    (2004)
  • J.A. Pruszynski et al.

    Optimal feedback control and the long-latency stretch response

    Exp. Brain Res.

    (2012)
  • J.Y. Nashed

    Influence of the behavioural goal and environmental obstacles on rapid feedback responses

    J. Neurophysiol.

    (2012)
  • E.J. Cheng et al.

    On the use of musculoskeletal models to interpret motor control strategies from performance data

    J. Neural Eng.

    (2008)
  • D.C. Knill

    Flexible, task-dependent use of sensory feedback to control hand movements

    J. Neurosci.

    (2011)
  • J. Diedrichsen et al.

    Reversal of bimanual feedback responses with changes in task goal

    J. Neurophysiol.

    (2009)
  • M. Dimitriou

    Task-dependent coordination of rapid bimanual motor responses

    J. Neurophysiol.

    (2012)
  • P.K. Mutha et al.

    Shared bimanual tasks elicit bimanual reflexes during movement

    J. Neurophysiol.

    (2009)
  • C.D. Manning

    Proprioceptive reaction times and long-latency reflexes in humans

    Exp. Brain Res.

    (2012)
  • L.H. Sergio et al.

    Hand and joint paths during reaching movements with and without vision

    Exp. Brain Res.

    (1998)
  • J. Izawa

    Motor adaptation as a process of reoptimization

    J. Neurosci.

    (2008)
  • J.A. Pruszynski

    Rapid motor responses are appropriately tuned to the metrics of a visuospatial task

    J. Neurophysiol.

    (2008)
  • J.A. Pruszynski

    The long-latency reflex is composed of at least two functionally independent processes

    J. Neurophysiol.

    (2011)
  • S. Rossignol

    Dynamic sensorimotor interactions in locomotion

    Physiol. Rev.

    (2006)
  • F. Crevecoeur

    Fast corrective responses are evoked by perturbations approaching the natural variability of posture and movement tasks

    J. Neurosphysiol.

    (2012)
  • L.P.J. Selen

    Deliberation in the motor system: reflex gains track evolving evidence leading to a decision

    J. Neurosci.

    (2012)
  • P.K. Mutha

    Visual modulation of proprioceptive reflexes during movement

    Brain Res.

    (2008)
  • L. Yang

    Rapid motor responses quickly integrate visuospatial task constraints

    Exp. Brain Res.

    (2011)
  • D.M. Wolpert et al.

    Motor control is decision-making

    Curr. Opin. Neurobiol.

    (2012)
  • P. Cisek

    Making decisions through a distributed consensus

    Curr. Opin. Neurobiol.

    (2012)
  • Cited by (0)

    View full text