Agreement Between Spatiotemporal Gait Parameters Measured by a Markerless Motion Capture System and Two Reference Systems—a Treadmill-Based Photoelectric Cell and High-Speed Video Analyses: Comparative Study (Preprint)

Abstract

BACKGROUND Markerless systems require no markers to be attached to the body, improving clinical feasibility and testing time. However, the lack of markers might affect the measurements’ accuracy OBJECTIVE This study aimed to determine the absolute reliability and concurrent validity of the Kinect™ system with the MotionMetrix™ software for spatiotemporal variables evaluation during running at comfortable velocity by comparing data with two widely used systems (i.e., OptoGait™ and high-speed video-analysis at 1,000 Hz) METHODS Twenty-five runners performed a running protocol on a treadmill at 12 km.h-1. The Kinect™-MotionMetrix™ combination reported spatiotemporal parameters during running (i.e. contact time [CT], flight time [FT], step frequency [SF] and step length [SL]) and were compared to two reference systems. The Kinect™-MotionMetrix™ system obtained higher coefficients of variation than the reference systems in every spatiotemporal parameter. RESULTS Regardless of the system, FT displayed the highest values (OptoGait™ 16.4%, video analysis 17.3% and MotionMetrix™ 23.2%). The rest of CVs reported were lower than 8.1%. The correlation analysis reported very large correlations (r>0.8, p<0.001) and almost a perfect association (ICC>0.81) between systems for all the spatiotemporal parameters, except for CT, which reported lower values. The Bland-Altman plots between systems revealed small systematic biases and random errors for SF and SL, but greater systematic biases and random errors for temporal parameters as compared with both OptoGait™ (CT: +3.0%, FT: -7.9%) and high-speed video analysis at 1,000 Hz (CT: +4.2%, FT: -11.3%). Accordingly, heteroscedasticity was found between systems for temporal parameters (r2>0.1). CONCLUSIONS Therefore, the results indicate that the Kinect™-MotionMetrix™ combination slightly overestimates CT and strongly underestimates FT according to both OptoGait™ system and high-speed video analysis at 1,000 Hz, but it is a valid tool to measure SF and SL according to those reference systems. The results also question the reliability of this system for determining temporal parameters.