Identification of amyotrophic lateral sclerosis using EMG signals

Abstract

Electromyography (EMG) signal is a biomedical signal that is used to measure the electrical current generated in muscles during neuromuscular activities. This paper presents a method which uses EMG signals collected from biceps brachii muscles to identify Amyotrophic Lateral Sclerosis (ALS) disease using machine learning method. The work presents basic analysis and classification of EMG signals recorded from healthy and ALS subjects for automated screening of ALS disease using signal processing. Some characteristics of EMG signals such as maximum amplitude and mean of amplitude are strategically quantified and classified by using neural network classifier to automate the diagnosis of ALS Disease. Experiments were carried on the EMG database created under EMG Lab, United States and results are encouraging. The experiment result indicates enough discrimination between ALS and healthy EMG signals for automated screening.