Automated EMG Signal Classification for Diagnosis of Neuromuscular Disorders Using DWT and Bagging

Abstract

Electromyographic (EMG) signals are used to diagnose the neuromuscular disorders. Machine learning algorithms are employed in the diagnosis of neuromuscular disorders as a decision support system. This paper present bagging ensemble classifier for the automated classification of EMG signals. Ensemble classifiers achieve better performance by using a weighted combination of several classifier models. Many researchers have shown the efficiencies of ensemble classifiers in different real-world problems. Though, there is almost no studies examining their feasibilities for diagnosis of neuromuscular disorders. Hence, in this study, the feasibility of bagging ensemble classifier is assessed for the diagnosis of neuromuscular disorders by using EMG signals. The method consists of three steps. In the first step, the Discrete Wavelet Transform (DWT) is employed for feature extraction from each type of EMG signals. Then, statistical values of DWT are calculated to represent the distribution of wavelet coefficients. In the last step, the obtained feature set is used as an input to a Bagging ensemble classifier for the diagnosis of neuromuscular disorders. Experimental results have shown the feasibilities of bagging ensemble classifier by achieving 99% accuracy with SVM for diagnosis of neuromuscular disorders.