Classification of EMG Signals Using an Efficient Hybrid Model (EMG + WPT + ANFIS + FFA) for the Detection of Neuromuscular Disorders

Electromyography (EMG) signals are produced for clinical application to detect the exact condition of muscular health. An EMG signal is developed to produce a relationship between man and machine to detect any disorder in muscular condition. Some parts of the body signals are extracted for detection, decomposition, and classification purposes. The basic purpose of this paper is to implement some of the efficient methods of decomposition, like the Wavelet Packet Tree (WPT), and classification, like the ANFIS methods, which are used to better understand the signal and its nature. This paper provides a comparative study also for researchers to get a better idea to implement EMG in clinical application. Our hybrid method (EMG+WPT+ANFIS) and (EMG+WPT+ANFIS+FFA) show statistical test results of sensitivity of about 97.63 and 98.81 and specificity of about 98.41 and 99.67.