Assessment Muscle Fatigue using Statistical Study & Classification: A Review Mohamed Sarillee, M.Hariharan, Anas M.N., Omar M.I, Aishah M.N. and Q.W.Oung Biomedical Electronic Engineering, School of Mechatronic Engineering, Universiti Malaysia Perlis Campus Pauh Putra, 02600 Arau, Perlis, Malaysia. sarilleelee@gmail.com Abstract—Muscle fatigue is very common muscle condition has been experienced. Different types of myograms have been used to assess muscle fatigue and there are Electromyogram, Mechanomyogram and Acoustic myogram. Each myogram has its own advantages in assessing muscle fatigue. Therefore, suitable features are needed to assess muscle fatigue. This review discusses the on statistical analysis and classification result of the myograms features that have been applied. Towards the end of the paper, challenges and future trends are discussed. Index Terms— Muscle fatigue, EMG, MMG, AMG, Statistical analysis I. INTRODUCTION Muscle fatigue is a very common muscle condition has been experienced in daily activity. Muscle fatigue occurs when the muscle could not maintain or perform expected forces or power [1, 2]. Assessment muscle fatigue is important in many fields such as sport sciences, clinical analysis and ergonomics. Therefore, different method and approach has been applied to assess muscle condition. The various techniques that have been used to assess the muscle condition are surface electromyography (sEMG), isometric strength tests, muscle biopsy (laboratory tests) and muscle imaging techniques. However, sEMG is commonly has been applied to assess the muscle condition by recording the muscle electrical activity which controls by the nervous system. The muscle contains motor unit action potential trains the action potential to adhesion motor unit in the muscle [3]. The sEMG is generated by summing up the action potential of the motor unit [4]. During fatigue the low frequency components and amplitude of the EMG signal increases due to recruitment of motor unit [5]. There are few limitations using sEMG such as ability to monitor only a few muscle sites, cross talk and electrode placement [6], which encourage to use another sensor such as accelerometer (vibration signal) and microphone (sound signal). Mechanomyogram (MMG) is mechanical micro vibration of muscle contraction. MMG technique provides information of muscle activities regarding muscle fiber change between fast or slow muscle fiber. The amplitude of the MMG is correlated with force production, even a small change in force which reflected in the MMG amplitude [7, 8]. Therefore, small changes in force during muscle fatigue were reflected in MMG amplitude, where the EMG amplitude does not follow the force production [7]. The advantages of MMG are, they easy and simple to implement. Further, MMG do not contain power line interference. They also have highest signal and noise ratio [9, 10]. That contracting muscle generates sounds has been known for nearly two centuries (Wollaston, 1810). The sound or acoustic signal called as acoustic myogram (AMG) is audible [11, 12]. The AMG is a low frequency signal, in 1810; William Hyde Wollaston discovered the frequency of the sound is 25 Hz [11, 13]. The nature of AMG was related to the pressure wave generated by the dimensional changes of the activated fibers, to the lateral movement of the muscle or to the molecular events of the contracting fibers. It has been proven that its amplitude and its frequency content are dependent on the contraction intensity. Assessment muscle fatigue has been quiet challenging for last two decades. The EMG signals have been used to assess muscle fatigue by applying different types of signal processing algorithm. In some studies, the MMG signals have been used to analysis muscle fatigue along with the EMG signals. The advantage of the MMG signals to reflect the force production very useful feature in assessing muscle fatigue. Furthermore, the researcher has been studying the relation between EMG and MMG signal features by using a statistical approach. In [11], AMG has been used to assess the muscle, the researcher state that the root mean square (RMS) of AMG reflected the load. The rest of the paper is organized as follows. Section II describes about analysis has been used to assess muscle fatigue. This follows by discussion in section III and last section (section IV) explain about conclusion and future work of the work. II. ANALYSIS A. Statistical Statistical analysis converts numbers into meaningful conclusions in accordance with the intentions of a survey. There are two major classes of statistical analysis. The first