[Computer-assisted evaluation of surface electromyograms of 8 different arm muscles]

The aim of the present study was, via digitization of the EMG signals and processing them in the computer, to make possible effective evaluation of electromyographically detectable modifications of the activity of the forearm muscles developing during wrist movements under defined conditions of loading. The search for the optimal measuring parameter was of particular importance. In groups of test subjects the electric voltage potentials generated by 8 arm muscles were measured simultaneously on the skin and evaluated using suitable variables (integral, RMS [root mean square] value, mean amplitude, summed amplitude, amplitude peak count [peaks], zero crossings) with the aid of a PC and a computer program specially developed for the task. On evaluation it was found that the parameters reacting sensitively mainly to amplitude size--integral, RMS value, mean amplitude height and summed amplitude--reflected, with gradually decreasing clarity, the activities of each muscle. The correlation among the first three parameters was almost linear. The parameters sensitive primarily to frequency modulations, such as peaks and number of zero crossings, yielded little information under the load applied. They did not adequately reflect the extension and flexion of the hand.