This document proposes a method for automated diagnosis of muscle diseases using electromyography (EMG) signals. It involves applying wavelet decomposition to EMG signals to extract features. A Hilbert transform is used to represent the EMG signal analytically, and features are calculated from the analytical signal. These features are input to a convolutional neural network (CNN) classifier to categorize the EMG signal and diagnose muscle diseases. Simulation software with MATLAB is used to test this process, with the goal of early and accurate detection of neuromuscular disorders.