To apply the multi-channel surface EMG technique of the facial muscles (Schumann et al. 2010) in investigations of facial nerve paresis and its therapy the reproducibility of the method has to be evaluated. Therefore the surface EMGs of 18 healthy volunteers (6 male, 12 female, mean age 24.5 years, SD +/­ 1.5) were monopolarly recorded from the facial muscles in a test-retest scenario (48-channel, 10–700 Hz, 3000 samples/s, resolution of 2,44 mV/bit) using small disc surface electrodes (diameter 4 mm, Ag/AgCl). The time lag between the first and the second EMG registration was several days. Volunteers were instructed to perform 31 defined facial movements. EMG curves were visually checked, quantified by power spectral analysis and normalised.

Results: 

Movement-dependent activation profiles of the facial muscles were very similar as demonstrated by pairs of box plots. The Spearman rank correlation coefficient - used to compare activation profiles between test and re-test - were 0.88 (median), 0.65 (minimum), 0.81 - 0.91 (lower - upper quartile), 0.94 (maximum). In comparison between the first and the second EMG examination, multivariate analysis of variance (ANOVA, repeated measures) showed only in 10 of 31 movements a significant (p<0.05) difference. All in all, the results demonstrate a sufficient reproducibility of the facial surface EMG. Variations were hardly due to electrode placement but more to the performance of facial movements. The EMG method can be used in follow-up examinations.