Decomposition and quantitative analysis of clinical electromyographic signals

https://doi.org/10.1016/S1350-4533(99)00064-8Get rights and content

Abstract

Procedures for the quantitative analysis of clinical electromyographic (EMG) signals detected simultaneously using selective or micro and non-selective or macro electrodes are presented. The procedures first involve the decomposition of the micro signals and then the quantitative analysis of the resulting motor unit action potential trains (MUAPTs) in conjunction with the associated macro signal. The decomposition procedures consist of a series of algorithms that are successively and iteratively applied to resolve a composite micro EMG signal into its constituent MUAPTs. The algorithms involve the detection of motor unit action potentials (MUAPs), MUAP clustering and supervised classification and they use shape and firing pattern information along with data dependent assignment criteria to obtain robust performance across a variety of EMG signals. The accuracy, extent and speed with which a set of 10 representative 20–30 s, concentric needle detected, micro signals could be decomposed are reported and discussed. The decomposition algorithms had a maximum and average error rate of 2.5% and 0.7%, respectively, on average assigned 88.7% of the detected MUAPs and took between 4 to 8 s. Quantitative analysis techniques involving average micro and macro MUAP shapes, the variability of micro MUAPs shapes and motor unit firing patterns are described and results obtained from analysis of the data set used to evaluate the decomposition algorithms are summarized and discussed.

Introduction

The quantitative analysis of the motor unit action potentials (MUAPs) of individual motor units detected using monopolar needle (MN), concentric needle (CN) or single fibre needle (SFN) electrodes can provide important information for the diagnosis and treatment of neuromuscular disorders [1], [2], [3], [4], [5], [6]. Statistics, based on a motor unit's average MUAP shape, are calculated from a representative sample of motor units, usually of size 20, from a muscle of interest to extract morphological information regarding the size and fibre distributions of its motor units. Common statistics used with regard to MN and CN MUAPs include duration, amplitude, number of phases and turns, area, amplitude-area ratio, spike duration and size index [1], [2], [3], [4], [5]. The variability or jitter, of the temporal intervals between significant peaks within SFN MUAPs is analyzed to assess the functional stability of a muscle's neuromuscular junctions, while the average number of significant peaks is used to estimate the density of motor unit fibres in close proximity to the detection-surface [6].

Traditionally, such MUAPs are often obtained using level or window triggering during slight muscle contraction. Recently however, several clinical EMG signal decomposition systems have been introduced [7], [8], [9], [10], [11], [12] and at least three are currently available on commercial clinical EMG systems and have published reference values [10], [11], [12], [13], [14], [15]. EMG signal decomposition resolves a composite EMG signal into its constituent motor unit action potential trains (MUAPTs). The resulting MUAPTs represent the individual activities of a population of concurrently active motor units. Consequently, analysis of the MUAPTs can provide information regarding motor unit firing patterns and the stability of the MUAPs within each MUAPT in addition to the standard morphological information based on statistics of the average MUAP shapes. The stability of MUAPs within a MUAPT, or jiggle, can provide information related to that obtained using SFN MUAPs [16].

The decomposition of an EMG signal is a difficult task. The characteristics of each signal depend on the type of electrode used, its position relative to the muscle, the level of contraction and the clinical state of a subject's neuromuscular system. EMG signal decomposition algorithms must therefore be capable of providing robust performance across signals with widely varying characteristics. A system of algorithms for signal decomposition will be described and results of the evaluation of their performance across a wide spectrum of EMG signals will be presented. A brief description of the overall decomposition system has been presented earlier [17] and detailed explanations and evaluations of several of its key aspects have also been previously reported [18], [19], [20], [21]. However, the entire system has not been previously described in detail nor has its performance been reported. In addition, methods of analyzing the stability of the MUAPs within a MUAPT and combining the results of the decomposition of signals detected with selective or micro electrodes with signals acquired simultaneously using non-selective or macro electrodes will be introduced.

Section snippets

Methods

Following are descriptions of the major components of the EMG signal decomposition and analysis system.

Results

Ten sets of EMG signals detected during slight to moderate levels of contraction, corresponding to approximately 10 to 25% of maximum voluntary contraction (MVC), of the extensor digitorm communis (EDC) muscle of normal subjects were used for evaluation and to present exemplary normative data. EMG signals were detected from 4 healthy subjects ranging in age from 25 to 72 years. The typically 20–30 s long micro and macro signals were detected using CN and overlying 1 cm by 3 cm surface

EMG signal decomposition

Robust performance is very important for clinical application. The decomposition system presented successfully decomposed a variety of EMG signals with up to over 130 MUAPs per second. Decompositions were consistently accurate and for each signal sufficient information was provided to accurately calculate the various quantitative EMG parameters of interest even though superimposed MUAPs were not resolved. Due to memory constraints the current DOS implementation spends significant time accessing

Summary

The presented system for decomposing and analysing EMG signals is capable of extracting useful clinical information from simultaneously acquired micro and macro signals. Based on a set of 10 signals used for evaluation it was demonstrated that CN-detected micro signals could be decomposed with sufficient accuracy and speed to provide clinically useful parameter values relating to detailed aspects of the structure and function of the motor units of a muscle. In conjunction with the micro signal

References (38)

  • P Barkhaus et al.

    Quantitative EMG in inflammatory myopathy

    Muscle Nerve

    (1990)
  • F Buchthal et al.

    The diagnostic yield of quantified electromyography and quantified muscle biopsy in neuromuscular disorders

    Muscle Nerve

    (1982)
  • D.B Sanders et al.

    AAEM Minimonograph #25 single-fiber electromyography

    Muscle Nerve

    (1996)
  • A Gerber et al.

    A new framework and computer program for quantitative EMG signal analysis

    IEEE Trans. Biomed. Eng.

    (1984)
  • G.H Loudon et al.

    New signal processing techniques for the decomposition of EMG signals

    Med. & Biol. Eng. & Comput.

    (1992)
  • M.H Hassoun et al.

    NNERVE: Neural network extraction of repetitive vectors for electromyography — Part I: Algorithm

    IEEE Trans. Biomed. Eng.

    (1994)
  • K McGill et al.

    Automatic decomposition of the clinical electromyogram

    IEEE Trans. Biomed. Eng.

    (1985)
  • S.D Nandedkar et al.

    Multi-motor unit action potential analysis (MMA)

    Muscle & Nerve

    (1995)
  • J.E Howard et al.

    Properties of motor unit action potentials with concentric and monopolar needle electrodes: ADEMG analysis

    Muscle & Nerve

    (1988)
  • Cited by (170)

    • Near-fiber electromyography

      2021, Clinical Neurophysiology
    View all citing articles on Scopus
    View full text