Artificial neural net algorithms in classifying electromyographic signals

Abstract

The aim of this work is to examine how Artificial Neural Nets (ANN) can be used as a computerized method for electromyographic diagnosis. For this reason, a number of well defined neuromuscular disorders have been selected, ie. Becker's Muscular Dystrophy (BMD), which is an X linked myopathy affecting boys characterized by progressive atrophy and loss of muscle fibers. Spinal Muscular Atrophy (SMA), is a hereditary disease of the motor neuron, with progressive loss of motor units and a continuous process of reinnervation of the remaining ones, resulting in new large units. Reinnervation and fiber grouping are also noted in the initial stages of Motor Neuron Disease (MND), the third disease under study. This is a disease affecting middle to old aged people, with progressive widespread loss of motor neurons, usually leading to death within 3 to 5 years. In the advanced stages of this disease large units also denervate.