Bipolar intramuscular wire electrodes and spectral analysis of the electromyographic signal have been used to measure fatigue in muscles that cannot be studied with surface electrodes. Intramuscular electrodes can detect a greater range of frequencies from muscle, obtain a less distorted signal, and are therefore felt to be more sensitive to detecting fatigue. To determine the reliability and sensitivity of electrode placement (with a fixed distance) for assessing muscle fatigue, we placed three intramuscular electrodes in and two surface electrodes on the biceps brachii of 30 healthy male subjects. With these electrodes, we devised eight configurations that were analyzed separately for reliability. Subjects performed four, 30-s isometric fatiguing contractions divided between two testing sessions. Mean and median frequency of the power density spectrum were plotted against time. Linear regression was performed to obtain slopes, which were used as indicators of fatigue. The bipolar surface electrode configuration displayed mean and median frequency intrasession and mean frequency intersession reliability for slope. All four bipolar fine-wire configurations had mean and median frequency intrasession reliability (P < or 0.05). Only three of the four bipolar fine-wire configurations approached mean frequency intersession reliability, and none fo the four displayed median frequency intersession reliability. the configuration with distal bipolar intramuscular electrodes placed 1 cm apart was the most reliable intramuscular technique. The bipolar fine-wire configuration studied showed a trend toward better reliability than monopolar fine-wire configurations. No intramuscular technique, however, was reliable enough for clinical use in the study of fatigue.

Normal values for initial median frequency (IMF) of the electromyographic (EMG) power-density spectrum must be determined before EMG spectral analysis can be used to evaluate clinically muscle fatigue. This study attempts to establish normal values in four muscles. Thirty-one healthy subjects performed isometric contractions of the biceps brachii, triceps, deltoid, and tibialis anterior muscles at 50% maximum voluntary contraction. Linear regression analysis was used to compute the IMF and the slope of the median frequency as it decayed with fatigue over 45 seconds. The IMF of the tibialis anterior (mean +/- standard deviation, 116 +/- 20 Hz) was significantly higher (p < 0.001) than that of the biceps (90 +/- 18 Hz), triceps (85 +/- 18 Hz), and deltoid (87 +/- 15 Hz). The deltoid and tibialis anterior had the steepest slopes. The IMF of all muscles was greater in men than in women, but gender did not affect the slope. This study attempted to establish normal values for IMF and slope in specific muscles. However, the range of normal values is so broad that it may preclude the clinical use of spectral analysis to evaluate muscle fatigue.

Spectral analysis of the electromyogram (EMG) signal is a useful tool for studying the complex phenomena of muscle fatigue. The power density spectrum of the myoelectric signal undergoes a compression towards low frequency as a function of time during a sustained contraction. This compression has been measured using the spectral parameter, median frequency. Normal median frequency values were established for the biceps, triceps and deltoid during a pilot study. The triceps has a lower median frequency slope with fatigue which suggests that this muscle may have more type I fibers (fatigue resistant) than the biceps and deltoids