MES classification using artificial neural networks and chaos theory

Investigations into the nonlinear dynamics of biological signals have revealed that many biological signals, once thought to be stochastic, are in fact chaotic. Poincare sections are used in the study of nonlinear dynamics for the analysis and characterization of attractors in chaotic signals. This paper outlines a method that classifies myoelectric signals (MESs) based on their Poincare sections with the aid of an artificial neural network. Results are presented that have immediate application to a five-state prosthetic control system which is based on biceps brachii action alone. A short discussion follows on its potential for automating myopathy diagnosis.