A multi-level approach to understanding upper limb function

Prog Brain Res. 2007;165:347-62. doi: 10.1016/S0079-6123(06)65021-8.


Here we describe a multi-level approach to study upper limb control. By using non-human primates we were able to examine several different levels of motor organization within the same individual including their voluntary behavior, musculoskeletal plant, and neural activity. This approach revealed several parallels in the global patterns of activity of upper arm muscles and neurons in primary motor cortex (M1). For example, during postural maintenance both arm muscles and arm-related M1 neurons exhibit a bias in torque-related activity towards whole-limb flexion and whole-limb extension torque. A similar bias could be reproduced with a mathematical model of muscle recruitment that minimized the effects of motor noise suggesting a common constraint for the population activation of muscles and cortical neurons. That said, M1 neurons were not merely "upper motor neurons" as they exhibited substantial context-dependency in torque-related activity compared to arm muscles. This flexible association with low-level processing is consistent with M1 having a pivotal role in an optimal feedback controller.

Publication types

  • Research Support, Non-U.S. Gov't
  • Review

MeSH terms

  • Animals
  • Humans
  • Motor Cortex / cytology
  • Motor Neurons / physiology
  • Movement / physiology*
  • Muscle, Skeletal / physiology
  • Neural Networks, Computer
  • Upper Extremity / physiology*