A Multi-Classifier Approach to MUAP Classification for Diagnosis of Neuromuscular Disorders

Abstract

The shapes and sounds of isolated motor unit action potentials (MUAPs) in an electromyographic (EMG) signal provide a significant source of information for diagnosis, treatment and management of neuromuscular disorders. These parameters can be analyzed qualitatively by an expert or quantitatively by using pattern recognition techniques. Due to the advantages of quantitative EMG method, developing robust automated MUAP classifiers have been explored and several systems have been developed for this purpose by now, but the accuracy of the existing methods is not high enough to be used in clinical environments. In this paper, a novel classification strategy based on ensemble of support vector machines (SVMs) classifiers in hybrid serial/parallel architecture is proposed to determine the class label (myopathic, neuropathic, or normal) for a given MUAP. The developed system employs both time domain and time-frequency domain features of the MUAPs extracted from an EMG signal using an EMG signal decomposition system. Different classification strategies including single classifier and multiple classifiers with several subsets of features were investigated. Experimental results using a set of real EMG signals showed robust performance of multi-classifier methods proposed here. Of the methods studied, the multi-classifier that uses multiple features sets and a combination of both trainable and nontrainable fusion techniques to aggregate base classifiers showed the best performance with average accuracy of 97% which is significantly higher than the average accuracy of single SVM-based classifier system (i.e., 88%).

An Embedded, Eight Channel, Noise Canceling, Wireless, Wearable sEMG Data Acquisition System With Adaptive Muscle Contraction Detection

“…EMG technology is used in a wide range of applications such as medical diagnosis and rehabilitation [12]- [15], sports science research, athlete monitoring [16], human machine interaction and gesture recognition [17]- [20]. Specifically, non-invasive EMG technology, i.e., surface electromyography (sEMG), has been recently used in research with these wearable systems [21].…”

A Real-Time Method for Decoding the Neural Drive to Muscles Using Single-Channel Intra-Muscular EMG Recordings

“…According to this procedure, the motor unit action potential (MUAP) trains are extracted from the EMG. 4 Applications of EMG decomposition are in diagnosing neuromuscular diseases based on morphological measures of MUAPs, 5 estimating MU architectural properties, physiological and anatomical studies of the neuromuscular system, 6 studies on MU control [7][8][9] and neural connectivity, 10,11 and neurological assessments.…”

Longitudinal tracking of physiological state with electromyographic signals.

“…Kamali et al (2014) , look at the shape of the MUAPs themselves in order to distinguish between healthy, neuropathic, and myopathic conditions [2]. After decomposing an EMG signal into a MUAP train (a sequence of MUAPs), they extract features describing the MUAP in both the time domain (rise time, duration, peak-to-peak amplitude) and time-frequency domain via discrete wavelet transform (sub-band coefficients mean absolute value, standard deviation).…”

“…Observing and diagnosing the changing conditions of various physiological states in humans with algorithmic approaches is an important research field, and one that could aid in advancements of diagnostic and clinical technologies for non-invasive observation of the neuromuscular system [1,2]. For example, NASAs Strategic Knowledge Gap (SKG) reports from LEAG (Lunar Exploration Analysis Group) [3] and MEPAG (Mars Exploration Program Analysis Group) [4] indicate broadly that "how to maintain peak human health and performance in dusty, high-radiation, partial gravity environments" is a matter of concern.…”

“…A more algorithmic approach is exemplified in the work of Kamali et al (2014), whose approach utilizes a variety of classification algorithms and wavelet-derived features which will be described in Chapter 2 [2]. It is appropriate here to reiterate the principle aim of this work to explore the development of methodology for algorithmically marking long-term changes in physiological state.…”