1. Introduction
Electromyography (EMG) is the study of muscle function through analysis of the electrical signals emanated during voluntary or involuntary muscle contractions. EMG provides noninvasive means for the study of muscular function during biofeedback training, activities of sports or daily living [1], [2], [3]. It is also useful in interpreting pathological states of the musculoskeletal systems [4], [5]. In particular, EMG offers valuable information concerning the timing of muscular activity and its relative intensity [6]. Intramuscular fine wire electrode and surface electrodes are two ways to collect EMG data. Intramuscular fine wire electrodes have been considered for their field selectivity and their reduced range of pick up areas [7], [8]. They have increased bandwidth, a more specific pick-up area, and the ability to record from deep muscles. Because myoelectric signals propagate through muscles and layers of tissue, EMG signals detected by surface recording electrodes may be erroneously interpreted as generated by muscle fibers within the sampling field of the electrode. This phenomenon is known as cross talk [9], [10], [11]. Due to this cross talk, it may be impossible to detect EMG of very deep muscle whereas intramuscular fine wire electrode will be the ideal choice. However, intramuscular fine wire electrodes require a needle for insertion into the muscle and may cause a significant pain. The discomfort of intramuscular fine wire electrodes can increase the tightness or spasticity in the muscles. Moreover, intramuscular fine wire electrodes are less repeatable as it is very difficult to place the needle in the same area of the muscle each time. On the other hand, the data acquisition through surface electrodes is a free pain method to collect EMG. Surface electrode are more reproducible, easier to apply, and more suitable for movement and gait analysis. EMG analysis requires various metrics, such as the mean frequency (MNF), median frequency (MDF), the root mean square (RMS), the rectified root mean square (RRMS) and the average rectified value (ARV). These metrics are used to understand physiological phenomena during muscular contractions. Muscular activities are accompanied with muscle fatigue, which is defined as the reduction of power output in EMG signal [12]. It is an important concept in EMG analysis because muscle fatigue is directly correlated to safety in long term-period works. The metrics, MNF, MDF, RMS, RRMS, and ARV, are often used to analyze the muscle fatigue. For instances, the MNF and MDF are considered as appropriate indicators of spectral shifts as the muscle fatigue progresses. The RMS and ARV indicate the muscle fatigue by showing the changes in the amplitudes of EMG. Some studies have shown that RMS and ARV metrics from surface EMG are not sufficient for providing information about a single muscle. However, they agreed with the fact that the relationship between surface and intramuscular EMG is closely related with the muscles and their contraction types [13]. Therefore, the relationship between surface and intramuscular EMG should consider the size and location of the selected muscle. In this study, the correlation of muscle fatigue metrics between surface EMG and intramuscular EMG will be examined during isometric contractions.