Techniques for characterizing elbow and shoulder mechanics and neural control in normal and disabled subjects

This paper describes a robotic manipulator and several analysis techniques which have been devised to characterize human arm biomechanics and to investigate the neural control of arm posture and movement. The importance of limb stiffness in understanding normal neural control strategies and in restoring arm function to individuals with quadriplegia resulting from spinal cord injury is discussed. The manipulator has the capacity to impose constant loads of greater than 225 N and to impose stochastic force perturbations with a bandwidth of 50 Hz. A modification of a standard single input-single output system identification technique to identify endpoint stiffness properties is discussed. The use of these experimental and analytical techniques to study whole arm posture and movement, shoulder joint posture and internal joint stability, and arm control in quadriplegic subjects is described.