Motor learning and prediction in a variable environment

doi:10.1016/S0959-4388(03)00038-2

Current Opinion in Neurobiology

Volume 13, Issue 2, April 2003, Pages 232-237

https://doi.org/10.1016/S0959-4388(03)00038-2 Get rights and content

Abstract

Traditional studies of motor learning and prediction have focused on how subjects perform a single task. Recent advances have been made in our understanding of motor learning and prediction by investigating the way we learn variable tasks, which change either predictably or unpredictably over time. Similarly, studies have examined how variability in our own movements affects motor learning.

Introduction

In everyday life we are required to move about in a changing and often unpredictable environment. Despite these variations we remain able to achieve our behavioural goals with apparent ease. Motor control researchers have recently shown increasing interest in understanding how we learn to control our movements and predict the consequences of our actions in predictably and unpredictably varying environments. Computer controlled virtual environments, usually including a robotic manipulandum for force feedback, are often employed in this research as they allow experimenters to precisely control the parameters of their subject’s mechanical and visual environment. When subjects are exposed to a new mechanical environment their movements are initially perturbed, but return to approximately their normal pattern after several hundred movements. Here, we review recent studies looking at motor learning and prediction.

Section snippets

Motor learning in an uncertain environment

Several studies have examined how we learn tasks whose parameters vary randomly over time. Takahashi et al. [1] asked subjects to make elbow flexion and extension movements against a viscous load, the strength of which was drawn randomly from a Gaussian distribution in each trial. In this randomly varying environment, subjects learned the average of the loads they experienced. Scheidt et al. [2^•] confirmed that this was also true for planar reaching movements in which a velocity-dependent force

Motor prediction

The ability to predict the future state of the motor system is thought to be essential for skilled movement because of the delays inherent in the sensorimotor system. Recent evidence has suggested that we predict the consequences of our motor commands and the behaviour of external objects to generate current estimates of the state of our body and the environment.

It has recently been shown that we are able to predict subtle variations in the dynamic state of our arms. Ariff et al. [16] hid

Motor learning and consolidation

Several studies have examined memory consolidation after learning motor tasks. Memory consolidation is the process by which memory representations become increasingly robust with the passage of time. In the hours after learning a dynamic motor task, such as movement in a force field, progressive memory consolidation takes place 21., 22.. Unlike the consolidation of perceptual skills and sequence learning 23.•, 24., 25., 26., this process does not require a period of sleep [27]. It was recently

Optimal control

Most movement tasks can be achieved using many different joint configurations, levels of co-contraction and so on. Several studies have sought to understand why certain motor patterns are preferred to others (stereotypy). These studies place motor learning within an optimal control framework, in which a task is associated with a cost, for example, the energy consumed or the time taken to complete the task. Planning or learning can be considered to be part of producing the movement that best

Conclusions

We have reviewed recent advances in understanding motor learning and prediction. Progress has been made in understanding the effects on movement of a mechanical environment that varies both predictably and unpredictably. Our ability to predict the consequences of our own motor commands and the behaviour of external objects is also being revealed in increasing detail. Although we have some partial answers the coming years should elucidate which motor tasks lead to competition in motor working

References and recommended reading

Papers of particular interest, published within the annual period of review, have been highlighted as:

•
of special interest
••
of outstanding interest

References (37)

M. Ghilardi et al.
Patterns of regional brain activation associated with different forms of motor learning
Brain Res.
(2000)
C.D. Takahashi et al.
Impedance control and internal model formation when reaching in a randomly varying dynamical environment
J. Neurophysiol.
(2001)
R.A. Scheidt et al.
Learning to move amid uncertainty
J. Neurophysiol.
(2001)
A.G. Witney et al.
The influence of previous experience on predictive motor control
Neuroreport
(2001)
A. Karniel et al.
Does the motor control system use multiple models and context switching to cope with a variable environment?
Exp. Brain Res.
(2002)
N. Bhushan et al.
Computational nature of human adaptive control during learning of reaching movements in force fields
Biol. Cybern.
(1999)
V. Wigmore et al.
Visuomotor rotations of varying size and direction compete for a single internal model in motor working memory
J. Exp. Psychol. Hum. Percept. Perform.
(2002)
K.A. Thoroughman et al.
Electromyographic correlates of learning an internal model of reaching movements
J. Neurosci.
(1999)
J.V. Cohn et al.
Reaching during virtual rotation: context specific compensations for expected coriolis forces
J. Neurophysiol.
(2000)
T. Wang et al.
Learning the dynamics of reaching movements results in the modification of arm impedance and long-latency perturbation responses
Biol. Cybern.
(2001)

R. Nezafat et al.

Long-term adaptation to dynamics of reaching movements: a PET study

Exp. Brain Res.

(2001)

J.B. Dingwell et al.

Manipulating objects with internal degrees of freedom: evidence for model-based control

J. Neurophysiol.

(2002)

D. Rancourt et al.

Stability in force-production tasks

J. Mot. Behav.

(2001)

E. Burdet et al.

The central nervous system stabilizes unstable dynamics by learning optimal impedance

Nature

(2001)

J.W. Krakauer et al.

Independent learning of internal models for kinematic and dynamic control of reaching

Nat. Neurosci.

(1999)

C. Tong et al.

Kinematics and dynamics are not represented independently in motor working memory: evidence from an interference study

J. Neurosci.

(2002)

G. Ariff et al.

A real-time state predictor in motor control: study of saccadic eye movements during unseen reaching movements

J. Neurosci.

(2002)

J.R. Flanagan et al.

The inertial anisotropy of the arm is accurately predicted during movement planning

J. Neurosci.

(2001)

Cited by (69)

Building an adaptive interface via unsupervised tracking of latent manifolds
2021, Neural Networks
In human–machine interfaces, decoder calibration is critical to enable an effective and seamless interaction with the machine. However, recalibration is often necessary as the decoder off-line predictive power does not generally imply ease-of-use, due to closed loop dynamics and user adaptation that cannot be accounted for during the calibration procedure. Here, we propose an adaptive interface that makes use of a non-linear autoencoder trained iteratively to perform online manifold identification and tracking, with the dual goal of reducing the need for interface recalibration and enhancing human–machine joint performance. Importantly, the proposed approach avoids interrupting the operation of the device and it neither relies on information about the state of the task, nor on the existence of a stable neural or movement manifold, allowing it to be applied in the earliest stages of interface operation, when the formation of new neural strategies is still on-going.
In order to more directly test the performance of our algorithm, we defined the autoencoder latent space as the control space of a body–machine interface. After an initial offline parameter tuning, we evaluated the performance of the adaptive interface versus that of a static decoder in approximating the evolving low-dimensional manifold of users simultaneously learning to perform reaching movements within the latent space. Results show that the adaptive approach increased the representational efficiency of the interface decoder. Concurrently, it significantly improved users’ task-related performance, indicating that the development of a more accurate internal model is encouraged by the online co-adaptation process.
White matter microstructure changes induced by motor skill learning utilizing a body machine interface
2014, NeuroImage
Citation Excerpt :
Performing complex motor skills is a fundamental component of ordinary human life. The ability to learn and modify motor skills is a requisite for adapting to an ever-changing environment (Davidson and Wolpert, 2003). Through practice, new motor skills are acquired and existing ones are continuously refined.
The purpose of this study is to identify white matter microstructure changes following bilateral upper extremity motor skill training to increase our understanding of learning-induced structural plasticity and enhance clinical strategies in physical rehabilitation. Eleven healthy subjects performed two visuo-spatial motor training tasks over 9 sessions (2–3 sessions per week). Subjects controlled a cursor with bilateral simultaneous movements of the shoulders and upper arms using a body machine interface. Before the start and within 2 days of the completion of training, whole brain diffusion tensor MR imaging data were acquired. Motor training increased fractional anisotropy (FA) values in the posterior and anterior limbs of the internal capsule, the corona radiata, and the body of the corpus callosum by 4.19% on average indicating white matter microstructure changes induced by activity-dependent modulation of axon number, axon diameter, or myelin thickness. These changes may underlie the functional reorganization associated with motor skill learning.
The primate cerebellum selectively encodes unexpected self-motion
2013, Current Biology
Citation Excerpt :
However, sensory prediction errors arise not only as a result of changes in the motor apparatus and environment (i.e., conditions that drive motor learning) but also whenever we experience externally produced sensory stimuli. If externally imposed stimulation is systematically paired with voluntary movement, motor learning occurs [8]. In contrast, when sensory stimulation is unexpected, the computation of sensory prediction errors effectively enables the brain to distinguish between the consequences of our self-generated actions (sensory reafference) and stimulation that is externally produced (sensory exafference).
The ability to distinguish sensory signals that register unexpected events (exafference) from those generated by voluntary actions (reafference) during self-motion is essential for accurate perception and behavior. The cerebellum is most commonly considered in relation to its contributions to the fine tuning of motor commands and sensorimotor calibration required for motor learning. During unexpected motion, however, the sensory prediction errors that drive motor learning potentially provide a neural basis for the computation underlying the distinction between reafference and exafference.
Recording from monkeys during voluntary and applied self-motion, we demonstrate that individual cerebellar output neurons encode an explicit and selective representation of unexpected self-motion by means of an elegant computation that cancels the reafferent sensory effects of self-generated movements. During voluntary self-motion, the sensory responses of neurons that robustly encode unexpected movement are canceled. Neurons with vestibular and proprioceptive responses to applied head and body movements are unresponsive when the same motion is self-generated. When sensory reafference and exafference are experienced simultaneously, individual neurons provide a precise estimate of the detailed time course of exafference.
These results provide an explicit solution to the longstanding problem of understanding mechanisms by which the brain anticipates the sensory consequences of our voluntary actions. Specifically, by revealing a striking computation of a sensory prediction error signal that effectively distinguishes between the sensory consequences of self-generated and externally produced actions, our findings overturn the conventional thinking that the sensory errors coded by the cerebellum principally contribute to the fine tuning of motor activity required for motor learning.
Music performance as knowledge acquisition: a review and preliminary conceptual framework
2024, Frontiers in Psychology
Motor adaptation and immediate retention to overground gait-slip perturbation training in people with chronic stroke: an experimental trial with a comparison group
2023, Frontiers in Sports and Active Living
Human Stiffness Perception and Learning in Interacting With Compliant Environments
2022, Frontiers in Neuroscience

View all citing articles on Scopus

View full text

Motor learning and prediction in a variable environment

Abstract

Introduction

Section snippets

Motor learning in an uncertain environment

Motor prediction

Motor learning and consolidation

Optimal control

Conclusions

References and recommended reading

Brain Res.

Impedance control and internal model formation when reaching in a randomly varying dynamical environment

J. Neurophysiol.

Learning to move amid uncertainty

J. Neurophysiol.

The influence of previous experience on predictive motor control

Neuroreport

Does the motor control system use multiple models and context switching to cope with a variable environment?

Exp. Brain Res.

Computational nature of human adaptive control during learning of reaching movements in force fields

Biol. Cybern.

Visuomotor rotations of varying size and direction compete for a single internal model in motor working memory

J. Exp. Psychol. Hum. Percept. Perform.

Electromyographic correlates of learning an internal model of reaching movements

J. Neurosci.

Reaching during virtual rotation: context specific compensations for expected coriolis forces

J. Neurophysiol.

Learning the dynamics of reaching movements results in the modification of arm impedance and long-latency perturbation responses

Biol. Cybern.

Long-term adaptation to dynamics of reaching movements: a PET study

Exp. Brain Res.

Manipulating objects with internal degrees of freedom: evidence for model-based control

J. Neurophysiol.

Stability in force-production tasks

J. Mot. Behav.

The central nervous system stabilizes unstable dynamics by learning optimal impedance

Nature

Independent learning of internal models for kinematic and dynamic control of reaching

Nat. Neurosci.

Kinematics and dynamics are not represented independently in motor working memory: evidence from an interference study

J. Neurosci.

A real-time state predictor in motor control: study of saccadic eye movements during unseen reaching movements

J. Neurosci.

The inertial anisotropy of the arm is accurately predicted during movement planning

J. Neurosci.