Journal of Rehabilitation Medicine 51-6 | Page 63

Validity and reliability of a spatiotemporal gait-analysis system narrow width of the carpet and requiring them to walk on and off the carpet (19). Several low-cost systems exist that use a single-camera setup, footswitches, ac- celerometers, gyroscopes, and inertial measurement units; however, results on the accuracy of spatial para- meters are inconsistent or absent and it is questionable whether some of these systems are reliable in persons with gait deviations (e.g. forefoot contact at initial contact) (19–30). A 2-dimensional spatiotemporal gait analysis sys- tem (SGAS) at relatively low-cost (approximately one-tenth of the price of a GAITRite® system) was developed that measures spatiotemporal gait parame- ters in the sagittal plane using a single camera placed perpendicular to the walkway. The camera can either be used in a stationary position, or moved manually along a parallel rail system to capture multiple strides during a single walk. The SGAS uses custom software for camera calibration and position and time measure- ment (31). Individuals walk unobtrusively while their gait is recorded with the camera. To our knowledge no camera system using this calibration method has been validated for assessing spatiotemporal gait parameters. The aim of this study was to establish the concurrent validity of the SGAS for assessing the spatiotemporal parameters of gait in healthy subjects under 4 different walking conditions: barefoot walking, shod walking, and to mimic gait deviations that may result from neu- rological or musculoskeletal disorders, toe walking and slow walking. Furthermore, the minimum number of footsteps needed to achieve reliable estimates of spa- tiotemporal gait parameters, inter-rater and intra-rater reliability, and measurement error were determined. METHODS Subjects A sample of 33 healthy adults (13 men, 20 women, mean age 43 years, standard deviation (SD) 12 years) were recruited from employees, their relatives, and visitors to our rehabilita- tion centre through flyers posted at the centre (see Table I for subject characteristics). Individuals were eligible if they were Fig. 1. Spatiotemporal gait analysis system (SGAS): camera positioned on a movable tripod and the computer screen with the software. over 18 years of age, could walk independently, and were free of musculoskeletal or neurological pathology. Data collection took place at the human movement laboratory of our centre. The study protocol was approved by the medical ethics committee of the Academic Medical Centre of the Amsterdam University Medical Centres (protocol number NL50002.018.14). All sub- jects provided written informed consent. Instrumentation The SGAS consists of a high-definition 2.2-megapixel camera with 50× optical zoom (f/1.8–4.2, 16:9), sampling at 50 Hz (Panasonic Corporation, Osaka, Japan). Custom software on a Windows computer with a high-definition multimedia interface (HDMI) frame grabber was used for camera calibration, video recording, and position and time assessment (TMSi, Oldenzaal, The Netherlands). The SGAS software is available open source (https://github.com/MvanBloemendaal/SGAS). The camera was levelled and positioned on a movable tripod (camera dolly). In this study, the camera was positioned at a height of 92 cm and the perpendicular distance from a 10-m long walkway equals Table I. Subject characteristics (n=33) Characteristics Mean (SD) Age, years Body height, cm Body mass, kg Leg length, cm Foot length, cm Comfortable gait speed, m/s 43.1 (12.0) 177.9 (10.5) 76.0 (14.2) 92.4 (7.1) 26.5 (1.7) 1.2 (0.2) Cadence, steps per min a Step width, cm a 457 108.8 (8.7) 68.2 (6.9) Gait cycle time, s a Stride length left leg, cm a Stride length right leg, cm a a 1.1 (0.1) 137.2 (13.6) 137.2 (14.0) At comfortable gait speed measured by the GAITRite® system. Fig. 2. Calibration of the camera for determining the projection matrix. J Rehabil Med 51, 2019