EMG analysis of facial muscles exercise using oral cavity rehabilitative device F. Ibrahim 1 , J. H. Chae 2 , N. Arifin 1 , N. M. Zarmani 1 , and J. Cho 2 1 Department of Biomedical Engineering, University of Malaya, Kuala Lumpur, Malaysia 2 Department of Biomedical Engineering, Inje University, Gimhae, Republic of Korea Abstract - This study presents the analysis of surface electromyography (sEMG) of facial muscles in different age categories. The analysis was evaluated when the subjects performed the gripping oral exercise device in a static contraction. The sEMG measurements were taken at four positions; Orbicularis Oris Superiors Left (OOSL) and Right (OOSR), Orbicularis Oris Inferiors Left (OOIL) and Right (OOIR) using TeleMyo2400T. There are only 4 female subjects participated in the study which is divided into 2 Groups: Groups 1 and 2 with mean age of 47.5 and 23 years old, respectively. The findings indicate that the mean and median frequencies of sEMG activity recorded from both left and right of OOS and OOI muscles showed decreasing trend from the beginning until the end of the static contraction sessions. The power spectral density of sEMG signal on those muscles for both groups is in the range of 19 to 45 Hz. I. INTRODUCTION Mean Frequency (MF) and Median Frequency (MdF) appear to be one of the useful parameter in evaluating fatigue exercise of facial muscle. The central tendencies of the electromyography (EMG) power spectrum can serve this purpose as a measurable parameter. It has been well known that the EMG power spectrum declines to a lower frequency domain during pro-longed static contraction [1]. This frequency decline can be interpreted as a sign of localized muscle fatigue and can be objectively measured by surface electromyography (sEMG) power spectrum statistics: the MF and MdF [2]. The mean power frequency is the average of all considerable frequencies. The median power frequency is the frequency where the spectrum is divided in two parts of equal power [3]. The development of localized muscle fatigue can be measured by the collected EMG sig-nal of the power spectrum [4-6]. The power spectrum analysis utilizes the frequency do- main processing for information content. The Fourier trans- form procedure is used to transform the time domain sEMG signals into the frequency domain data [7]. With a static muscle contraction, the low frequency components of the signal gradually increase while the high frequency components decrease. It was suggested the combination of synchronization and desynchronization of motor unit firing as a possible cause of frequency shift [8]. The spectral density of the sEMG signal will give the power carried by the signal, usually per unit frequency. This is then known as the Power Spectral Density (PSD) of the signal. The sEMG signal detected during a muscle contraction is a random, non-stationary signal. It should be noted that when estimating the PSD of a random signal, the muscle is assumed to be stationary over the analysis time window. Thus, the EMG signal is effectively considered quasi-stationary, which makes the selection of the epoch length an issue that requires careful investigation. The energy content is the mean value of energy spectral density on time do-main. The energy spectral density is referred to PSD. This study proposes four different types of signal processing techniques to analyze the of sEMG of OOS and OOI in a fatiguing static contraction such MF, MdF, PSD and energy contents in different age categories. II. SUBJECTS AND METHODS A. Subjects There are only 4 female subjects participated in the study which are divided into 2 groups according to their age category. Groups 1 and 2 consist of mean age of 47.5 and 23 years old, respectively. The subjects are healthy without functional disturbances or pain complaints in jaw, face, and head region, and with natural dentition. They were given an oral rehabilitative device (Patakara® LIP Trainer) and were asked to grip the oral device in a static contraction. B. Measurement of EMG signal Alcohol swap was used to removes makeup and dirt for a clean area of suction and the sEMG measurement were conducted in the sitting position. The main criterion for determining the recording sites was to maximize the distance to adjacent muscles (to reduce crosstalk), the locations of the innervations zones could not be considered in electrode positioning, because the facial motor units have neuromuscular junctions distributed over the whole muscle in several oval-shaped zones. The electrodes are positioned as close as possible to the vermillion border (due to the size of the double-adhesive tape used for attachment, there was a minimal distance of 1cm between the electrode center and the vermillion border). The medial electrodes are 18mm from the facial midline and then mark 8 positions on the face of subject [9]. After mark the positions, attach the electrodes which are cut to small size (10mm) and then connect to pre-amplifier wires, fixed as surgical tape. The subject is observed them- selves in a recorded monitor from video camera; the monitor 1-4244-0549-1/06/$20.00 ©2006 IEEE.