فهرست نشریات علمی دسترسی آزاد ایران

Electromyography (EMG) signal analysis plays a crucial role in the diagnosis and monitoring of various neuromuscular disorders, including myopathy and neuropathy. This study proposes a robust pattern recognition framework for classifying healthy, myopathic, and neuropathic EMG signals using a set of computationally efficient time-domain features. Pre-processed EMG signals were segmented into 1-second windows, from which Root Mean Square (RMS), Mean Absolute Value (MAV), Zero Crossings (ZC), Waveform Length (WL), Skewness, and Kurtosis were extracted. These features were then used to train and evaluate three popular machine learning classifiers: Support Vector Machine (SVM), Random Forest, and K-Nearest Neighbors (k-NN). The results demonstrate high classification accuracy, with SVM and k-NN achieving 100% accuracy, and Random Forest achieving 97.83%. This comparative study highlights the effectiveness of time-domain features combined with traditional machine learning algorithms for robust EMG pattern recognition in clinical applications, offering a viable alternative or complement to more complex deep learning approaches, especially in scenarios with limited diverse datasets.