Decomposition and Analysis of Intramuscular Electromyographic Signals

The clinical community has long shown interest in the concept of extracting as many motor unit action potentials (MUAPs) as possible from an intramuscular electromyographic (EMG) signal. Adrian and Bronk (1929) developed the first concentric needle electrode to identify both shape and firing rate of the MUAPs. Subsequent manual approaches of graphically measuring and quantifying the EMG signal evolved into computer-based techniques directed at identifying individual action potentials and discharge times by shape discrimination. The Precision Decomposition technique described in this chapter recovers all the usable information available in the EMG signal. The information can be conveniently grouped into two categories: morphology and control properties. Morphology describes the parameters of the MUAP shape such as the peak-to-peak amplitude, the time duration, the number of phases, and the area. These parameters are provided by the recovered Concentric and Macro MUAP. The morphology of the MUAP describes features that are related to the anatomical and physiological properties of the muscle fibers. These are the parameters which the clinician is accustomed to evaluating during a standard clinical EMG examination. The control properties of the motor units dictate the firing characteristics of the motor units. Therefore, the firing characteristics provide a description of how the motor units are controlled by the central nervous system and to some extent the peripheral nervous system. Clinically, they quantify upper motoneuron diseases. The technique of Precision Decomposition has been under development by our group since the late 1970s. The first public description of it was in the form of an abstract published in the Abstracts of the Society for Neuroscience (LeFever and De Luca, 1978). The signal processing concepts which underlie the approach appeared in the IEEE Transaction of Biomedical Engineering (LeFever et al. 1982a, b). A more pragmatic description of the algorithms and workings of the technique was provided by Mambrito and De Luca (1984). This paper also described a generic foolproof method of measuring the accuracy of any decomposition technique. Stashuk and De Luca (1989) have provided an update on useful modifications and applications of the technique while De Luca (1993) recently provided a comprehensive, hands-on account of the methodology.