Gait and Posture 20 (2004) 102���107 Technical note Evaluation of an ambulatory system for gait analysis in hip osteoarthritis and after total hip replacement K. Aminian a,���, C. Trevisan b, B. Najafi a, H. Dejnabadi a, C. Frigo c, E. Pavan c, A. Telonio c, F. Cerati b, E.C. Marinoni b, Ph. Robert a, P.-F. Leyvraz a,d a Center for Orthopaedic Research, Laboratory of Movement Analysis and Measurement, Swiss Federal Institute of Technology (EPFL), Bio-E, Bat. ELG, CH-1015 Lausanne, Switzerland b Department of Orthopaedic Surgery, University of Milano Bicocca, San Gerardo Hospital, Monza, Italy c Centro di Bioingegneria, Fnd. Don Gnocchi ONLUS-IRCCS, Politecnico di Milano, Milan, Italy d H�� opital Orthop��dique de la Suisse Romande, Lausanne, Switzerland Received 22 November 2002 received in revised form 21 May 2003 accepted 16 June 2003 Abstract Spatial and temporal parameters of gait have clinical relevance in the assessment of motor pathologies, particularly in orthopaedics. A new gait analysis system is proposed which consists of (a) an ambulatory device (Physilog��) including a set of miniature gyroscopes and a portable datalogger, and (b) an algorithm for gait analysis. The aim of this study was the validation of this system, for accuracy and clinical applicability. Eleven patients with coxarthrosis, eight patients with total hip arthroplasty and nine control subjects were studied using this portable system and also a reference motion analyzer and force plate. The small differences in the stance period (19 �� 31 ms), stride length and velocity (0.4 �� 9.6 cm and 2.5 �� 8.3 cm/s, respectively), as well as thigh and shank rotations (2.4 �� 4.3��� and 0.3 �� 3.3���, respectively), confirmed good agreement of the proposed system with the reference system. In addition, nearly the same accuracy was obtained for all three groups. Gait analysis based on Physilog�� was also in agreement with their Harris Hip Scores (HHS): the subjects with lower scores had a greater limp, a slower walking speed and a shorter stride. This ambulatory gait analysis system provides an easy, reproducible and objective method of quantifying changes in gait after joint replacement surgery for coxarthrosis. �� 2003 Elsevier B.V. All rights reserved. Keywords: Gait analysis Ambulatory system Hip osteoarthritis Gyroscope Force plate 1. Introduction Gait analysis [1] can be used for evaluating improve- ments after an intervention to improve walking. Compre- hensive gait analysis usually includes kinematics, kinetics and electromyography [2] and this complex information can only be obtained in a dedicated laboratory. Kinematics, kinetics and electromyography are fundamental to charac- terize gait patterns and their underlying mechanisms [3,4]. However, simplified analysis using spatio-temporal param- eters, for example, or comparisons between populations can also be valuable clinically and a portable device may be advantageous for these types of applications. Usually, portable devices only provide a few measurements, typi- ��� Corresponding author. Tel.: +41-21-693-2617 fax: +41-21-693-2614. E-mail address: kamiar.aminian@epfl.ch (K. Aminian). cally spatio-temporal using foot switches [5] or joint angles from electrogoniometers [6]. They have a limited accuracy that depends on the measuring device itself and also on how they are attached to the patient. However, they have the advantage of not requiring a gait analysis laboratory, are easy to use even by non-expert operators and provide results quickly, which are easy to interpret. One potential application of a portable device for gait analysis could be in the field of hip or knee arthroplasty. Although functional scores are used widely to assess change the patients��� responses are often subjective and the disparity between the patient���s and doctor���s evaluations significant [7,8]. Similarly, radiological signs do not always correlate with functional outcomes of arthroplasty [9,10]. Therefore, objective, dynamic and quantified data from gait analysis could be of additional interest to integrate with other clinical and instrumental data. A portable system may expand the clinical use of gait analysis. 0966-6362/$ ��� see front matter �� 2003 Elsevier B.V. All rights reserved. doi:10.1016/S0966-6362(03)00093-6

K. Aminian et al. / Gait and Posture 20 (2004) 102���107 103 In this study, a new gait analysis system is proposed which consists of (a) an ambulatory device (Physilog��) including a set of miniature gyroscopes and a portable datalogger, and (b) an algorithm for gait analysis. The aim of this study is to show the applicability of this ambulatory system as a com- plementary method of gait analysis after hip arthroplasty. 2. Methods 2.1. Gait analysis using a force plate and an ELITE system Gait Analysis was performed using a motion analyzer (ELITE system, BTS SrL, Milan, Italy) equipped with four video cameras recording at 100 Hz [11,12]. One 60 cm �� 40 cm force plate (Kistler 9661 A) was embedded in the floor about one-third of the way down a 10 m walkway. Signals from the forceplate were collected at 500 Hz. The acquisi- tion protocol has already been described [13]. The retrore- flective markers were positioned as follows: fifth metatarsal head, lateral malleolus, lateral femoral epicondyle, poste- rior iliac spines and lower edge of sacrum, the posterior process of the seventh cervical vertebra and the point of maximum thoracic kyphosis. Joint kinematics were con- fined to flexion/extension angles at the hip, knee, and ankle joints. The spatio-temporal parameters were computed from the initial and terminal contacts of each foot on the ground. The force platform was used to detect the initial contact of one foot and was defined when the ground reaction force achieved a threshold of 5% of body weight. The next contact of the foot on the ground occurred in a non-instrumented area of the floor. Contact was defined when a similar markers��� configuration was achieved (same distance from the floor, same displacement) as in the contact detected by the force platform. This method has been described previously [14]. The contralateral side was analyzed by evaluating the dis- tance of the foot from the ground. 2.2. Gait analysis using Physilog�� system Physilog�� (BioAGM, CH) is an ambulatory system based on miniature kinematic sensors attached on body segments and a portable recorder placed on the waist belt. The de- vice is lightweight (300 g), compact, can record up to eight channels, allows the storage of up to 16MB of data on a re- movable memory card and can operate continuously for up to 16 h on a rechargeable battery. It thus offers a practical method of gait analysis during daily activities. In this study, lower limb movement during walking was measured us- ing four miniature gyroscopes (Murata, ENC-03J) attached, respectively, to each shank and each thigh (Fig. 1). Each sensor measured the angular rotation rate parallel to the medio-lateral axis. The signals were digitized (12 bit) at a sampling rate of 200 Hz by the portable data logger, and Fig. 1. A subject carrying the Physilog�� system. The datalogger (300 g) can be fixed around the waist. Each sensor (gyroscope) is attached with elastic strip on each shank and thigh and linked with thin cable to the datalogger. stored for off-line analysis. An original method based on wavelet analysis was proposed to compute the values of spa- tial and temporal gait parameters from the angular velocity of the lower limbs [15]. First, gait phases were determined from the precise moments of left initial contact (ICL), left terminal contact (TCL), right initial contact (ICR) and right terminal contact (TCR). Every temporal parameter of a gait cycle, k, was computed as a percentage of this gait cycle as shown in Fig. 2. These parameters were the duration of stance, swing, initial double stance (IDS), terminal double stance (TDS), the sum of initial and terminal double stance corresponding to double support (DS) and the absolute dif- ference of initial and terminal double stance (ADDS). Spa- tial parameters were estimated by integrating the angular rate of rotations of the thigh and shank. The maximum and minimum of each angle during each cycle was detected. The difference between the maximum and minimum angles was considered as the range of rotation and based on these val- ues the following angular parameters were computed and compared to the ELITE system: ��� range of rotation for left and right thigh: R��L, R��R ��� range of rotation for left and right shank: R��L, R��R ��� range of rotation around left and right knee: R�� L, R�� R. In addition, the stride length (SL) and stride velocity (SV) were divided by the height of each subject for normalization purposes.