Decomposition of intramuscular EMG signals using a heuristic fuzzy expert system

Although increasingly sophisticated algorithms have been proposed to decompose intramuscular electromyography signals into the concurrent activities of individual motor units (MUs), the human operator is still able to improve decomposition results by visual inspection. The rationale for this paper was to combine components from previous decomposition procedures in an expert systems approach utilizing fuzzy logic and attempting to replicate the thought process of an accomplished decomposer in order to minimize the user interaction subsequently needed to enhance decomposition results. The decomposition procedure is discussed and examples are given of the type of information it can yield. The method has been used to identify the discharge activities of up to 15 MUs with up to 95% accuracy.