A neural network model of force control based on the size principle of motor unit

Abstract

A neural network model consisting of single motor cortex output cell, a-motoneurons, Renshaw cells and muscle units is presented. Firing rates of amotoneurons at different levels of muscle force obtained from the model closely agree with those measured in human muscles. Isometric force of the model increases almost linearly with the discharge frequencies of motor cortex output cell, in much the same fashion as being observed in monkeys. Effects of the size and the number of a-motoneurons in a motoneuron pool and inhomogeneous inhibitory actions via Renshaw cells upon behaviors of motor units are investigated.

Adaptive neural networks for control of movement trajectories invariant under speed and force rescaling

“…Without it, increments in P produce diminishing returns of force output from each muscle in the system. With it, force development is approximately linear in P. Compatible results regarding only a-MN, Renshaw cell interactions may be found in a one-muscle-channel simulation study of force output by Akazawa, Kato, and Fujii (1989).…”

Classification of motor commands using a modified self-organising feature map

“…In Section I, Data 1 was the fire rate of the motor units which again was approximately proportional to the muscle tension [33,34], thus Data 1 in the human set-up correspond to Data 1 in Section I. Eq. (1) from Section I was used to generate Data 2, and for data set Data 3 the amplitude was added (2), increasing the number of inputs to the network by one element.…”

“…The biological data were, therefore pre-processed to simplify the task for the ANNs. Muscle tension depends on spike frequency [33,34]. Counting spikes during a fixed time frame is thus an appropriate first step in pre-processing the data, and it is a standard method of intramuscular EMG quantification [35].…”

“…The EMG spike count vector would, therefore, when a certain movement is recorded, only differ in length and not in direction, because the spike frequency is proportional to the force [33,34]. The second set of data (Data 2) is pre-processed with Eq.…”

An oscillator theory of motor unit recruitment

“…Motor neurons exhibit a range in the size of their cell bodies and hence in the magnitude of their activation thresholds. Thus as the excitatory command signal grows in steps, the number of motor neurons that respond also increases in a graded manner resulting in a fine variation of muscle force (Akazawa and Kato 1990). The size principle mainly addresses the gain control issue in transforming a command stimulus into a graded force, but it seems to overlook the issue of synchronization (or desynchronization) among motor units.…”