Power spectrum analysis of the EMG pattern in normal and diseased muscles

The diagnostic value of analogue frequency analysis of EMG from patients with neuromuscular disorders has not been convincing. Using fast Fourier transformation it is today possible to obtain the EMG power spectrum on-line and with a better resolution. We examined the power spectrum of the EMG pattern of the brachial biceps muscle in 20 control subjects, 20 patients with myopathy, and 12 with neurogenic disorders. The electrical activity was sampled with a concentric needle electrode from 10 sites in each muscle. From each spectrum, mean power frequency, the power at 140 Hz, 1400 Hz, 2800 Hz and 4200 Hz relative to total power and the high/low ratio (1400/140) were obtained. The mean power frequency was higher at 10% than at 30% of maximal force. At a force of 30% of maximum the power spectrum analysis identified 55% and 64% of the patients with myopathy and neurogenic disorders, respectively. Although the diagnostic yield of the power spectrum analysis at a force of 10% of maximum was less than that at 30%, additional patients were identified at 10% increasing the diagnostic yield to 65% and 73% for patients with myopathy and neurogenic disorders, respectively. The best diagnostic parameters were the mean power frequency and the relative power at 1400 Hz.