Race walking is an event dictated by a rule that states that no visible loss of contact with the ground should occur and that the leg must be straightened from first contact with the ground until the ‘vertical upright position’ (IAAF Rule 230.2). During competition, compliance with the rule is assessed subjectively by judges but during biomechanical testing it is important to measure the knee angle objectively and accurately. The aim of this study was to compare the measurement of knee angles between 2D video and 3D optoelectronic systems during race walking. Seven elite male race walkers (stature: 1.77 m (± 0.03), mass: 65.7 kg (± 6.2)) and six elite female race walkers (stature: 1.66 m (± 0.08), mass: 58.6 kg (± 9.1)) participated in the study; in total, eight had competed at the Olympic Games. 2D video data were collected at 100 Hz using a high-speed camera. A 12-camera 3D optoelectronic motion capture system (Qualisys) operating at 250 Hz simultaneously captured the motion of three lower leg markers. The video files were digitised in two ways: first, through manually digitising by a single experienced operator; and second, using SIMI Motion’s automatic tracking function to track the three retroreflective markers. The optoelectronic files were processed through Qualisys Track Manager. All sets of knee angle data were filtered using residual analysis and interpolated to 101 points using a cubic spline. The root mean square difference (RMSD) between conditions was calculated for each individual, averaged across their five trials, and then averaged across all participants. The RMSD between the two visual digitising methods for one gait cycle was 3° (± 1). The RMSD between manual digitising and Qualisys was 4° (± 1), whereas between automatic tracking and Qualisys it was 2° (± 1). At initial contact, the mean angle calculated using manual digitising was 181° (± 2), using automatic tracking 180° (± 4), and using Qualisys 180° (± 4). The maximum angle during midstance was 185° (± 4) using manual digitising, 183° (± 5) using tracking, and 183° (± 5) using Qualisys. The minimum angle during midswing was 100° (± 6) using manual digitising, 102° (± 5) using tracking, and 101° (± 5) using Qualisys. Overall, all three methodologies gave similar results with no difference greater than 2° at any discrete gait event. It was unsurprising that the automatic tracking function in SIMI Motion and Qualisys produced similar knee angles given they used the same three joint markers, and showed that any movement of the athletes’ lower limbs out of the sagittal plane had little effect on joint angle calculation. In practical terms, using a markerless, 2D video system gave similar results to using a 3D optoelectronic system, meaning that it is appropriate for analysing in competition.
ORCID: 
CORE (COnnecting REpositories)
CORE (COnnecting REpositories)