An approach to identify myopathy disease using different signal processing features with comparison

Abstract

Myopathy, one of the most frequent inherited musculoskeletal diseases resulting in muscular weakness. Muscle cramps, tautness & spasm are also associated with myopathy. The electromyography (EMG) signals are biomedical signals that examine the muscle function through the inquiry of the electrical signal the muscles emanate. As the nervous system controls the muscle activity, the EMG signals can be viewed and analyzed in order to recognize the indispensable features of myopathy disease in individuals. The aim of this work is to dissociate the myopathic signals by studying the time & frequency domain features of the EMG signals. In this paper, autocorrelation (ACR), zero crossing rate (ZCR) as time domain features, mean frequency as frequency domain feature and Short Time Fourier Transform (STFT) as Time-frequency feature; are extensively analyzed on EMG signals of both the normal persons and the patients to successfully distinguish the patients from normal group. In order to comprehend this aim, EMG signal database was obtained from a normal control group consisted of 6 healthy persons & a group of patients with myopathy consisted of 6 patients. The analytical results show that myopathic signals have lower autocorrelation peak then the healthy ones and zero crossing rate and mean frequency of the affected signals are higher than the normal persons. In addition to that the power level of the spectrogram of the myopathy patients is considerably lower than that of the normal group and frequency shifting to higher frequency region for peak values.