EMG-BASED FATIGUE MEASUREMENTS DURING A DYNAMIC HIGH-INTENSITY EXERCISE

Changes in electromyogram (EMG) median frequency during isometric contractions have been used extensively to quantify muscle fatigue in human subjects. Unfortunately, the application of this technique to dynamic high-intensity activities may result in misleading fatigue measurements. This is due, in part, to muscle temperature changes and the documented influence of these changes on the EMG signal. The purpose of this study was to establish an EMG-temperature compensation technique that could correct for muscle temperature changes during the performance of dynamic high-intensity activities. In phase one of this study, the relationship between muscle temperature and EMG median frequency was determined for the Vastus Lateralis muscle (VL). Subjects were 4 healthy males between the ages of 27–42. The muscle was heated with diathermy to a temperature of approximately 40°C. Isometric EMG data were collected during the performance of a non-fatiguing protocol as the muscle cooled (39–34°C). The results revealed a linear relationship between EMG median frequency and muscle temperature (n = 7, mean slope = 2.85 ± 60 Hz/degree Celsius, R2 > .79). This was consistent with previous works. The mean slope was used as a muscle temperature correction factor that quantified the influence of muscle temperature on EMG median frequency. In phase two of this study, the subjects rode a lower extremity ergometer at workloads of 25%, 50%, and 75% of their VO2 max. Dynamic EMG data was collected as the subjects exercised. In addition, isometric EMG and VL intramuscular temperature data were collected at 10-minute intervals. The muscle temperature correction factor established in phase one was used with the measured change in muscle temperature to compensate for the effects of temperature on the EMG median frequency. To date, the successful use of this temperature correction technique with isometric muscle contractions has been demonstrated with some consistency. Efforts are now directed toward improving the performance of this method and extending the temperature correction methodology for use with dynamic EMG data.