Machine Learning in Neuromuscular Disease Classification

Abstract

This work reviews the recent techniques used to analyze the EMG signals to extract and classify features. The discussed techniques are used in medical applications for the diagnosing of neuromuscular disorders. Techniques such as wavelet transform (WT), principal component analysis (PCA), empirical mode decomposition (EMD), artificial neural network (ANN), support vector machine (SVM), extreme learning machine (ELM), k-nearest neighbors (kNN), and deep learning are among the powerful techniques used for feature extraction and disease classification. The many improvements made to the algorithms could increase the performance and accuracy of the models discussed in this area. Recently, the researchers focused their attention on the classification of neuromuscular illnesses using artificial intelligence to acquire and classify myoelectric signals via electromyography (EMG). Amyotrophic lateral sclerosis (ALS) is one of the neuromuscular illnesses that drives the attention of the researchers in this field. The current work focuses on the most recent methods for extracting and categorizing characteristics from EMG data to diagnose neuromuscular disorders in medical applications. Feature extraction approaches presented here include the wavelet transform (WT), principal component analysis (PCA), and empirical mode decomposition (EMD). Artificial neural networks include artificial neural networks (ANNs), support vector machines (SVM), extreme learning machines (ELM), k-nearest neighbors (kNN), and deep learning. An example on the uncertainty evaluation in machine learning classification is introduced.KeywordsEMG signalDisorder classificationMachine learningUncertainty