USC Inscription
Prof. Francisco Valero-Cuevas
Brain-Body Dynamics Lab
Please visit our new website at http://valerolab.org

Myths and Monsters in Motor Control

21st Annual Conference of the Society for the Neural Control of Movement

Friday, April 29, 2011, 8:00 am - 10:15 am

Chairs: Francisco J Valero-Cuevas

Participants: Jason J Kutch, Gerald Loeb, and Evangelos Theodorou

Abstract: Progress in the field of Sensorimotor Neuroscience hinges on a clear and up-to-date understanding of its fundamental tenets. While the validity of these tenets is routinely discussed and debated in the literature, it is also necessary to present them to the community in a manner that challenges all of us to affirm, update or revise them. This is particularly true when challenges or revisions to these tenets hinge on the details of specialized computational approaches, a nuanced interpretation of multiple experimental findings, a thorough knowledge of a vast literature, integration of recent findings, and/or careful semantic distinctions. We have assembled a team of four investigators with relevant expertise to critically review four fundamental tenets in light of recent work:

1) Muscle redundancy; Valero-Cuevas

2) Henneman's Size Principle; Kutch

3) Spinal Circuitry for Motor Behavior; Loeb

4) Optimal Control; Theodorou

The Panel will progress systematically from the periphery inwards to update and revise our understanding of these four critical aspect of Sensorimotor Neuroscience. We begin by discussing how the physics of tendon-driven systems defines the problem to be solved by the nervous system - and go on to show that we, in fact, have barely enough muscles for versatile behavior. While these concepts do not change the fact that there exist multiple solutions for "simple" tasks (i.e., with few constraints) and in the close vicinity of a valid coordination pattern for a "complex" task (i.e., with multiple constraints), they do strongly set the scope of feasible solutions, necessitate "co-contraction" and "synergistic" patterns independently of neural considerations - and make the system vulnerable to loss or weakness of even single muscles.

Next we transition to a reassessment of the physiological machinery that drives the recruitment of motor units - and challenge the notion that recruitment system is as autonomous and robust as is generally believed. A noninvasive innovative approach called EMG-weighted averaging can reveal the profile the motor unit activation pattern across the voluntary range (0-100% MVC) of an isometric contraction. By linking this experimental approach with computational models of motor unit activation we find that motor unit recruitment is itself tuned in accordance with their direction of mechanical force output.

Third, a computational assessment of the most thorough compendium of spinal circuitry to date sheds light on the high-level central drive necessary and sufficient to produce complex voluntary movement - and suggest the system can robustly meet these demands in any given individual with minimal tuning of the parameters of the circuitry.

The final part discusses the limitations of the optimal control framework and its implications to neuromuscular systems. Optimal control requires accurate knowledge of the dynamics and cost functions. Even though these requirements may be satisfied in tasks such as small reaching movements, they are strongly violated in learning scenarios or control tasks in which the dynamics and rewards are not known or vary (i.e., the majority of motor behaviors). We will provide an overview of other control theoretic and machine learning approaches - seldom applied in our field - motivated by these limitations and well suited to address them. These include reinforcement learning and model predictive, robust, and adaptive control.

Supplemental Material

1) Valero-Cuevas: Muscle redundancy

Presentation slides (in pdf)
References:

  • Kutch JJ., and Valero-Cuevas FJ.
    Muscle redundancy does not imply robustness to muscle dysfunction
    Journal of Biomechanics, 2011 Apr 29;44(7):1264-70. Epub 2011 Mar 21
    Additional Location
  • Valero-Cuevas FJ.
    A mathematical approach to the mechanical capabilities of limbs and fingers.
    Adv Exp Med Biol 629: 619-633, 2009. (Book Chapter)
  • Valero-Cuevas FJ, Venkadesan M, Todorov E.
    Structured variability of muscle activations supports the minimal intervention principle of motor control (Cover Article).
    Journal of Neurophysiology 102: 59-68, 2009.
    Supplemental material

    Back Cover image Front Cover image

  • Valero-Cuevas FJ.
    An integrative approach to the biomechanical function and neuromuscular control of the fingers.
    Journal of Biomechanics, 2005:38(4): p.673-84.
    ASB Post-Doctoral Young Scientist Award paper.
  • Jonathan L. Pearlman, Stephanie S. Roach, and Francisco J. Valero-Cuevas.
    The fundamental thumb-tip force vectors produced by the muscles of the thumb.
    Journal of Orthopaedic Research. 2004 22:306-312.
  • Valero-Cuevas FJ, Zajac FE and Burgar CB.
    Large index-fingertip forces are produced by subject-independent patterns of muscle excitation.
    J Biomech. 1998 31(8):693-703.
    *Related papers
  • 2) Kutch: Rethinking motor unit activation in health and disease (Henneman's Size Principle)

    Presentation slides (in pdf)
  • Muscle Activation Animation - Quicktime plug-in must be installed.
  • MATLAB code for EMG-weighted averaging, including instructions for use, example data, and accomanying publication.
  • 3) Loeb: Spinal Circuitry for Motor Behavior

    Presentation slides (in pdf)

    References:

  • Giby Raphael, George A. Tsianos, and Gerald E. Loeb.
    Spinal-Like Regulator Facilitates Control of a Two-Degree-of-Freedom Wrist.
    The Journal of Neuroscience 30(28):9431-9444 2010.
  • Cheng, E.J. and Loeb, G.E.
    On the Use of Musculoskeletal Models to Interpret Motor Control Strategies from Performance Data.
    Journal Neural Engineering 5:232-253, 2008.
  • Loeb, G.E.
    Learning From the Spinal Cord.
    J. Physiol. (London) 533:111-117, 2001.
  • Loeb, G.E.
    Overcomplete Musculature or Underspecified Tasks?
    Motor Control, 4:81-83, 2000.
  • Loeb, G.E., Brown, I.E. and Cheng, E.
    A hierarchical foundation for models of sensorimotor control.
    Exp. Brain Res., 126:1-18, 1999.
  • He, J., Levine, W.S. and Loeb, G.E.
    Feedback gains for correcting small perturbations to standing posture.
    IEEE Trans on Automatic Cointrol, 36:322-332, 1991.
  • Loeb, G.E., Levine, W.S. and He, J.
    Understanding sensorimotor feedback through optimal control.
    Cold Spring Harbor Symp. Quant. Biol. 55: 791-803, 1990.
  • 4) Theodorou: Is Optimal Control a Panacea?

    Presentation slides (in pdf)
    page last modified on November 20, 2015.
    currently maintained by: emily lawrence
    originally designed by: kornelius rácz