Series on Biomechanics, Vol.31, No.4 (2017), 3-11 3 Kinematic and vibration description of running pattern using empirical mode decomposition X. Chiementin a , M. Munera b ,C. B. Machado c , E. Abdi d , D. Sá-Caputo e , M. Bernardo-Filho f , R.Taiar a a GRESPI, Moulin de la Housse, Université de Reims Champagne Ardenne, 51687 Reims Cedex 2; France b Escuela Colombiana de Ingeniería Julio Garavito, Bogotá D.C., Colombia c Biomedical Ultrasound Lab., Estácio de Sá University, R. Eduardo Luiz Gomes,134,Niterói, Rio de Janeiro, Brasil d Montclair State University, Upper Montclair, New Jersey, USA e Departamento de Biofísica e Biometria, Universidade do Estado do Rio de Janeiro, Rio de Janeiro, RJ, Brazil. f Laboratório de Vibrações Mecânicas e Práticas Integrativas e Complementares, Universidade do Estado do Rio de Janeiro, 20551-030, Rio de Janeiro, RJ, Brasil _____________________________________________________________________________________ Abstract Background: Running is the source of repeated shocks which can reach amplitude over 20g during a foot-ground contact generating stress vibration in humans. This study proposes to apply an advanced signal processing method which is the Empirical Modal Decomposition (EMD) to describe the vibration at the tibia after the impact in running. Methods: One participant completed 5 tests at 10 and 14 km/h on a treadmill using Rearfoot strike and Forefoot strike techniques. Accelerometric data were collected at a sampling frequency of 1344 Hz. The sports kinematics and vibrations generated by the impact were dissociated by a reconstruction scheme of Intrinsic Mode Functions (IMF). Then, Root Mean Square (RMS) values were computed before and after the EMD. Results: EMD method allows (i) dissociating three phases generated by the runner kinematics and by the foot/ground shock, (ii) to estimate an RMS value generated to the impact (RearFoot, 14km/h: 6m/s², 10km/h: 4.5m/s²; ForeFoot, 14km/h:11.2m/s², 10km/h: 8.1m/s²). This value depends on the running technique and on the speed. Conclusion: This study is orientated on the original methodology of the signal treatment. Like perspective we will focus on the assessment between the findings of the current work with the publications of other studies. Keywords: Empirical modal decomposition (emd), running, vibratory risk, signal treatment, biomechanics __________________________________________________________________________________________ 1. Introduction Long distance running can increase the risk of degenerative disorders in lower extremities, particularly the knee joint [1]. The injuries are not caused by running itself, rather they are caused by either a structural weakness which the runner was born with, a postural weakness developed, or other stress caused by shoes or surfaces with poor shock attenuation capacities [2, 3]. In running, repeated impacts increase the risk of an overuse injury. That is why, to prevent injuries, many studies use the measure of acceleration, in particular at the tibia [4]. The components frequencies of the tibial acceleration for different pattern were described [5]. Other authors highlight that the surface has no effect on the tibial acceleration magnitude i.e. 10.3 to 12.4g. [6]. The numbers of studies using accelerometric analysis in running have increased over the past years. The interest in this subject is related to the emergence of Inertial Measurement Unit (IMU) in indoor and outdoor activities [7]. The IMUs are widely used attributable to good characteristics regarding the constraints of space, lightness and autonomy imposed for the measurement of human activities [8,9,10]. These devices have been deployed to evaluate the kinematics of athletes. However, their development have also allowed to conceive its use in the evaluation of vibratory risk [11,12]. Using this kind of sensors, acceleration have been measured at the tibia in running and walking to