Abstract—Sport injuries can prevent athletes from maintaining a high level of performance; keep them away from competitions during long periods of recovery, resulting in loss of income or even career. An objective system to assess the nature of the sport injury, adopt effective rehabilitation training at each stage and determine more effectively the time that the injured athletes can safely return to competition can reduce such liabilities. The diagnosis and monitoring of rehabilitation of sport injuries are critical for professional athletes. This study explores the feasibility of using inertial sensors to objectively aid in the assessment of sporting injuries and to monitor the progress made by an injured athlete during rehabilitation. This is proposed through observation, and contrasting the resultant deficits of muscle strength, and balance control, between healthy and injured parts of the body. The posture and gait of an athlete is monitored using a motion capture suit consisting of 17 tri-axial inertial sensor units. Three methods of Centre of Mass (COM), Symmetry angle (SA) and Gaussian Mixture Models (GMM) are applied to injured and control limbs, for a group from professional athletes when they perform single leg stance and one meter forward hopping. All of the three methods successfully distinguish between injured and control limbs. The results are presented and future extension of the methods is discussed. I. INTRODUCTION A. Ankle Injury Ankle injuries are one of the most common lower extremity injuries [1], and among such 80% are ligamentous sprains [2]. The total cost in healing ankle sprain reaches approximately 3.8 billion dollars per year in the world [3]. In spite of such high cost, up to 73% of individual ankle sprains have residual symptoms, repeated sprains and episodes of ‘giving way’ as a result of the non-full recovery [4]. The condition of Chronic Ankle Instability (CAI) can be further developed by repetitive ‘giving way’ in functional and mechanical deficits [5]. For a professional player, the unwelcomed yet common sport injury can result in loss of income or job [6]. A club manager *Resrach supported by St George Illawarra Dragons Rugby team. Shuo Zhang is a Ph.D candidate in School of Electrical, Computer & Telecommunications Engineering, University of Wollongong (e-mail: sz740@uowmail.edu.au ). Fazel Naghdy is now the Director of Centre for Intelligent Mechatronics Systems, University of Wollongong, Wollongong, NSW 2500 AU (e-mail: fazel@uow.edu.au). David Stirling is with the Centre for Intelligent Mechatronics Systems, University of Wollongong, Wollongong, NSW 2500 AU (e-mail: stirling@uow.edu.au). Montserrat Ros is with the Centre for Intelligent Mechatronics Systems, University of Wollongong, Wollongong, NSW 2500 AU (e-mail: montse@uow.edu.au). Andrew Gray is the Performance Director of St George Illawarra Dragons Rugby team (e-mail: agray@dragons.com.au) can lose a player at a vital time or face an unexpected cost to replace the injured player. Hence, it is critical to administer an appropriate treatment and rehabilitation processes for such injured athletes. Another challenge is to decide when injured athletes can return to competition. An early return may increase the risk of receiving more severe and even chronic injuries, whereas a late return can be financial penalizing for both the team and the athlete. B. Motion capture Developing an objective method to the diagnosis, prevention, and rehabilitation stages of sport injuries has been pursued by a number of research groups. They have compared the results of clinical assessment of injured and non-injured lower extremities or athletes using inertial and optical sensors combined with 3D modeling. Ramirez et al. [7] deploy force plate data in the posterior-medial direction and an inertial sensor on the spine. These sensors provide acceleration data along the z axis and wavelet transform along the y axis on anterior posterior movement to objectively assess postural excursion balance tests on injured athletes with CAI. Brown and Mynark utilize acceleration and EMG signals to estimate and monitor the uncomfortable tibia nerve stimulation during a double leg balancing stance tests to standardize an external perturbation [8]. Chu et al. [9] use a force plate to acquire the vertical ground reaction force, and the VICON optical motion tracking system to measure ankle inversion angle and velocity for ankle sprain injury prevention. Based on the results obtained, they suggest a threshold for the ankle inversion velocity. Velocities above this threshold they suggest will lead to ankle sprain injury. C. Primary Aim This study explores the feasibility of objectively measuring the progressive rehabilitation of injured athletes by monitoring the motion and posture of such using a motion capture system consisting of an array of inertial wireless sensors. With pattern recognition and advanced signal processing techniques, the motion of an injured athlete was monitored and characterized regularly during recovery. The data derived from two trials: single leg stance and one meter forward hopping are analyzed by three methods, the Centre of Mass (COM), Symmetry angle (SA) and Gaussian Mixture Models (GMM). This leads to a better understanding of an athlete’s condition. According to this result, team physiotherapists and performance directors can better determine the degree of an athlete’s injuries and develop a plan to further targeting strengthening exercises and optimize the recovery. It is expected that the high cost and the rate of Ankle Injury Assessment using Inertial 3D Data Shuo Zhang, Fazel Naghdy, David Stirling, Montserrat Ros and Andrew Gray Published in: Shuo Zhang; Naghdy, F.; Stirling, D.; Ros, M.; Gray, A., "Ankle injury assessment using inertial 3D data," IEEE/ASME International Conference on Advanced Intelligent Mechatronics (AIM), 2013, vol., no., pp.810, 815, 9-12 July 2013, doi: 10.1109/AIM.2013.6584193