Elsevier

Neural Networks

Volume 9, Issue 8, November 1996, Pages 1265-1279
Neural Networks

Forward Models for Physiological Motor Control

https://doi.org/10.1016/S0893-6080(96)00035-4Get rights and content

Abstract

Based on theoretical and computational studies it has been suggested that the central nervous system (CNS) internally simulates the behaviour of the motor system in planning, control and learning. Such an internal “forward” model is a representation of the motor system that uses the current state of the motor system and motor command to predict the next state. We will outline the uses of such internal models for solving several fundamental computational problems in motor control and then review the evidence for their existence and use by the CNS. Finally we speculate how the location of an internal model within the CNS may be identified. Copyright © 1996 Elsevier Science Ltd.

Section snippets

BACKGROUND

The topic for this section of the special issue of Neural Networks is whether the CNS makes use of internal models. Although shown to be of potential use in motor control, and finding applications in fields such as robotics, neural network and adaptive control, until recently there had been little evidence for control strategies used in man that are based on internal models. There is now growing support amongst researchers in human motor control that model-based strategies are used in the

Cancelling Sensory Reafference

A forward model is a key ingredient in a system that uses motor outflow (also called efference copy: Sperry, 1950; Festinger and Cannon, 1965; Kelso, 1977) to anticipate and cancel the sensory effects of movement. Sensory signals arise in the periphery from two causes: those as a result of environmental influences on the body, and those resulting from self-generated movement. The first are termed afference, while the second type of sensory signals are known as reafference as they are the

LEARNING AND REPRESENTATION OF INTERNAL MODELS

In this section we focus on two attributes of forward models: adaptability and representation. As we grow many of the parameters of the motor system, such as link lengths and inertias, which govern the dynamics of the motor system change dramatically. Hence, a forward model which captures the dynamics of a three-year-old child's arm is unlikely to be of use to the fully grown adult. Similarly, on a shorter time scale the dynamics of the motor system change when we pick up an object or even

EVIDENCE FOR FORWARD MODELS

We now turn to the evidence for forward models. As before we intend to restrict ourselves mainly to considering the human control of arm movement, but of course, we will also draw on animal studies providing electrophysiological evidence supporting internal models.

LOCALISATION OF A FORWARD MODEL

We now turn to the question of where these forward models may be found. Of course, it will be clear from the previous sections that forward models could be used within a number of motor systems, in a number of different ways. There might be several forward models in different brain sites. For example, a forward model has been proposed for ocularmotor control, and would be expected to be in brain stem circuits; a forward model used for high level motor planning is more likely to be found in

PHYSIOLOGICAL IDENTIFICATION OF AN INTERNAL MODEL IN THE BRAIN

In this final section, we ask how a forward (or inverse) model might be detected electrophysiologically, or by some physiological interventions. It is perhaps best to split this problem into several parts.

CONCLUSIONS

The evidence for internal models in physiological motor systems is still indirect. However, there are a number of experiments that point to the existence and use of internal forward models. Of course, many problems remain to be resolved.

Where in the brain are these models held? Localisation of internal models will likely come about through investigation such as neuronal recording, stimulation, lesioning, and functional imaging. We believe that several such lines of evidence point to the

Acknowledgements

RCM is supported by a Wellcome Senior Research Fellowship. We would like to acknowledge the support of the Wellcome Trust; we also thank the McDonnell-Pew Foundation and the MRC for support.

References (72)

  • C.G. Atkeson

    Learning arm kinematics and dynamics

    Annual Review of Neuroscience

    (1989)
  • F.L. Bedford

    Constraints on learning new mappings between perceptual dimensions

    Journal of Experimental Psychology

    (1989)
  • D. Bullock et al.

    A self-organising neural model of motor equivalent reaching and tool use by a multijoint arm

    Journal of Cognitive Neuroscience

    (1993)
  • F.W.J. Cody et al.

    Proprioceptive guidance of human voluntary wrist movements studied using muscle vibration

    Journal of Physiology (London)

    (1990)
  • Cody, F.W.J., Lovgreen, B., & Schady, W. (1991). Human visuo-motor tracking in cerebellar dysfunction. Journal of...
  • Craig, J.J. (1986). Introduction to robotics. Reading, MA:...
  • H. Cruse et al.

    Solution of the direct and inverse kinematic problems by a common algorithm based on the main multiple computations

    Biological Cybernetics

    (1993)
  • C. Dugas et al.

    Responses of cerebellar Purkinje cells to slip of a hand-held object

    Journal of Neurophysiology

    (1992)
  • E. Espinoza et al.

    Purkinje cell simple spike activity during grasping and lifting objects of different textures and weights

    Journal of Neurophysiology

    (1990)
  • L. Festinger et al.

    Information about spatial location based on knowledge about efference

    Physiological Review

    (1965)
  • J.R. Flanagan et al.

    Modulation of grip force with load force during point-to-point arm movements

    Experimental Brain Research

    (1993)
  • R.S. Gellman et al.

    Inferior olivary neurones in the awake cat: Detection of contact and passive body displacement

    Journal of Neurophysiology

    (1985)
  • Ghahramani, Z., Wolpert, D.M., & Jordan, M.I. (1995). Computational structure of coordinate transformations: A...
  • C. Ghez et al.

    Roles of proprioceptive input in the programming of arm trajectories

    Cold Spring Harbor Symposium on Quantum Biology

    (1990)
  • C. Ghez et al.

    Impairments of reaching movements in patients without proprioception. II. Effects of visual information on accuracy

    Journal of Neurophysiology

    (1995)
  • H. Gomi et al.

    Adaptive feedback control models of the vestibulocerebellum and spinocerebellum

    Biological Cybernetics

    (1992)
  • Goodwin, G.C., & Sin, K.S. (1984). Adaptive filtering prediction and control. Englewood Cliffs, NJ:...
  • J. Gordon et al.

    Impairments of reaching movements in patients without proprioception. I. Spatial errors

    Journal of Neurophysiology

    (1995)
  • P. Haggard et al.

    Coordinate responses following mechanical perturbation of the arm during prehension

    Experimental Brain Research

    (1995)
  • P.N. Haggard et al.

    Damage to cerebellocortical pathways following closed head injury: an MRI and behavioural study

    Journal of Neurology, Neurosurgery and Psychiatry

    (1995)
  • C. Hall et al.

    Imagery and the acquisition of motor skills

    Canadian Journal of Sports Science

    (1992)
  • C.S. Harris

    Perceptual adaptation to inverted, reversed and displaced vision

    Psychological Review

    (1965)
  • J. Hermsdorfer et al.

    Perturbation of precision grip in Friedreich's ataxia and late-onset cerebellar ataxia

    Movement Disorders

    (1994)
  • Hertz, J., Krogh, A., & Palmer R.G. (1991). Introduction to the theory of neural computation. Redwood City, CA: Addison...
  • H. Imamizu et al.

    Internal representations of the motor apparatus: Implications from the generalization in visuo-motor learning

    Journal of Experimental Psychology

    (1995)
  • R.A. Jacobs et al.

    Adaptive mixture of local experts

    Neural Computation

    (1991)
  • Cited by (1605)

    View all citing articles on Scopus
    View full text