Comparison of the Reproducibility of EMG Signals Recorded from Human Masseter and Lateral Pterygoid Muscles

EMG recordings of the left and right masseter and lateral pterygoid muscles were repeated three times in 15 young adults to test the reproducibility of the signals. Two tests were made on the same day (morning and afternoon) and one test three days later (morning). Needle and hook electrodes were used for the masseter muscle and needle electrodes for the lateral pterygoid muscle. The superficial areas of recording were tattooed on the skin. An intra-oral approach was selected for the lateral pterygoid muscle. The subjects were asked to perform three definite tasks for the masseter muscle (intercuspid occlusion, maximum protrusion, and centric relation with a bite opening of 3mm) and two for the pterygoid muscle (maximum opening and maximum protrusion). The EMG signals were directly integrated, and root mean squares of the AC components were computed. The selection of the signals to discard possible artefacts was performed by displaying EMG and RMS outputs on a six-channel UV oscillograph. The selected RMS values were submitted to different analyses of variance to define additive and non-additive models of effects. The results may be summarized as follows: The between-day sessions showed more variation than the within-day sessions. The variations related to the method accounted for a very high percentage of the total variance (48.9%). The recordings with needle electrodes displayed a large percentage of the individual variation (masseter, 31.2%; lateral pterygoid muscle, 65.9%) and significantly lower values for the variations related to the session or the side of recording. The one-side recordings were more suitable for distinguishing the biological variations than were the bilateral ones. The right side was less sensitive to variations than the left side. The study of the lateral pterygoid muscle was less dependent on "session", "side", or "task" effects than was that of the masseter muscle. The lack of significant variation resulting from the performed tasks suggested that only the lower bundle was recorded with a similar activity in protrusion and opening. The reproducibility of the superficial recordings of the masseter muscle may be improved if non-significant interactions between "task" and "area" or "side" factors, as well as between "task" and "session", are taken into consideration. A direct comparison of the different tasks performed in one "session" will help reduce the undesirable effects due to the random selection of the recording "side" or "site".