Neural network control of simple limb movements

Abstract

It is possible to embed the control and computation of a simple single-joint movement at different speeds by a small non-linear network of neuron-like elements. The network "learns" by appropriate adjustment of the strengths of interconnection, or synaptic weights, between the neuron-like elements. The learning of a few movement trajectories is generalized to the learning of a family of unlearned trajectories. These observations are in support of our hypothesis that relaxation of a network from an initial state to a final equilibrium state is both causal and computational to movement generation and control.

Strategies for the control of voluntary movements in patients with Parkinson's disease

“…To find some essential properties of a model with the help of computer simulations, one has to execute billions of iterations. This, in particular, may explain an opinion expressed in an earlier paper by Kwan et al (1990): "Different movement strategies (Gottlieb et al 1989) can be embedded in higher dimensional vector fields with multiple basing attractor structures. "…”

“…The task of the neural network is to synthesize a nonlinear dynamical system within which a movement strategy is embedded implicitly (Kwan et al 1990). We proceeded with the synthesis in the following way: First, we represented the time evolution of the three variables associated with one of the movement speeds as a temporally discrete sequence of 3vectors: a 0 , a L .…”

“…The ensuing relaxation describes an orbit ( Figure 1A) and generates dynamical behaviors ( Figure IB, solid trace), which are in excellent agreement with experimental data ( Figure 5 of target article). The learning of just a few sample orbits by a nonlinear network therefore leads to the creation of a family of potentially useful orbits (Kwan et al 1990), a process somewhat akin to transfer of motor learning. Learning more sample orbits will lead to a finer sculpturing of the basin structure and a correspondingly finer pattern generation, or control of movement.…”

Parkinsonian Tremor and Simplification in Network Dynamics

“…The model presented here, though somewhat abstracted from the details of connections of the basal ganglia and related structures, responds to this need. A number of recent network models of the basal ganglia and motor function (Kwan et al, 1990;Mitchell et al, 1991;Borrett et al, 1993;Contreras-Vidal and Stelmach, 1995;Suri et al, 1997) have concentrated on voluntary movement and learning, whereas we tackle an involuntary movement for the first time from a network perspective. In this context, too, we take the approach that basal ganglia functions are understandable as part of a larger motor network.…”

Cognitive Microgenesis

“…From this basic thermodynamic consideration, microgenesis is an energy-consuming irreversible process that must involve specific patterns of brain "work" or "activation" involved in the structuring of specific system instabilities. The appearance of the overt act itself may actually represent a "relaxation" of the potential state generated through the micro genetic process, with associated dissipation of energy (Kwan, 1988;Kwan, Yeap, Jiang, & Borrett, 1990). The microgenetic process of "setting up" the performance, however, would require the active selection of boundary conditions and the structuring of the singular set in the state space of the dynamics being prepared.…”