Issn Print: 0195-9131

Publication Date: 1996/05/01


Print

A WITHIN-SUBJECT SELECTION CRITERION FOR EMG DATA 457

G. W. Lange; R. A. Hintermeister

    loading  Checking for direct PDF access through Ovid

Excerpt

Technical problems that may affect EMG signal characteristics over time for surface (SE) and intramuscular fine wire electrodes (IWE) have been discussed in the literature. Movement of IWE within the muscle is a particularly difficult problem to control. To insure the quality of EMG data, our lab has checked the integrity of EMG electrodes by testing for differences in pre- and post-test MVC's across subjects. However, we have noted that unacceptable intraindividual differences may be masked by interindividual differences in such an analysis. For this study, we applied a within-subject criterion requiring the mean post-MVC to be within two standard deviations of the mean pre-MVC. Peak EMG amplitude (mv) of five standardized MVC's from the subscapularis (IWE) and the trapezius (SE) were recorded (2000Hz) from 15 subjects before and after a protocol of shoulder exercises lasting about one hour. The raw data were processed using a RMS algorithm with a 15ms moving window.
Although, repeated measures ANOVA indicated no differences between pre- and post-test MVC's, applying the selection criterion reduced the acceptable sample size by two subjects for the trapezius, and by eight subjects for the subscapularis. Intraclass reliability increased from 0.957 to 0.987 for the trapezius, and from 0.655 to 0.936 for the subscapularis. Correspondingly, the percentage of variance due to differences in mean pre- and post-MVC's decreased from 6.3% to 0.8% for the trapezius, and from 41.8% to 1.8% for the subscapularis. These data suggest that a within-subject acceptance criterion is appropriate to insure that the EMG electrode configuration is providing a consistent measurement of muscle activity over the course of data collection.
    loading  Loading Related Articles