Computer-Assisted Decomposition of the Electromyograms

The paper analyses the scope of multimodal parametric search (MPS) as the means for the decomposition of the electromyograms (EMGs). The approach is based on a predeened set of signal sources that are described by some parameters. These signal sources are searched for in a superimposed signal by calculation of the Euclidian distance and looking for its minimum. The signal sources are applied successively, according to the change of their parameters, and peeled oo at the minimum Euclidian distance found. The minima detected are then displayed in parametric space as a parametric representation. We showed that such a technique is reliable enough for the decomposition of needle EMGs at low contraction forces. Applied to surface EMGs, it still discerns diierent phenomena, although not always with a direct relationship to the reality. It is because of the method's ambiguity, amplitude ratio dependency, and a bias in the distribution of the detected signal components. Finally, we introduce a novel approach based on the wavelet transform and using the elements of the parametric search. Applied to the EMGs, both give comparable decomposition results. Electrical activity of muscles has been known and measured for a long time as electromyograms (EMGs). The measurements may be done invasively by needle electrodes and non-invasively by surface electrodes. So far, the EMGs were successfully involved in a variety of physiological analyses: in estimation of conduction velocities in the motor unit (MUs) 1], in nding out the location of innervation zones, in tracking the eeects of muscular fatigue 6], in estimation of the depth of certain MUs 7], in mathematical modelling of a muscle response 5], in detection of abrupt changes of muscle activity 2], and in some preliminary attempts of the EMG decomposition 9, 4]. Especially the decomposition of the EMG signals plays an important role in the diagnosing procedures. However, it can be done automatically, by computer-assisted approaches, accurate and reliable enough, more or less, only for needle EMGs with low contraction forces. All other possibilities, in particular in conjunction with surface recordings, is a topic of intensive investigation of several expert groups worldwide. Our research group has worked on some new techniques in order to extract the basic building blocks of the EMG signals, i.e. the motor-unit potentials (MUPs), in voluntary contractions. The rst attempts were devoted to low-force needle EMGs. Later on, after obtaining some good decomposition results on artiicial signals produced by our EMG …