Assessment of spatiotemporal gait parameters using a deep learning algorithm-based markerless motion capture system

Spatiotemporal parameters can characterize the gait patterns of individuals, allowing assessment their health status and detection clinically meaningful changes in gait. Video-based markerless motion capture is a user-friendly, inexpensive, widely applicable technology that could reduce barriers to measuring spatiotemporal clinical more diverse settings. Two studies were performed determine whether measured using demonstrate concurrent validity with those marker-based pressure-sensitive mat. For first study, thirty healthy young adults treadmill at self-selected speeds while synchronized video data recorded simultaneously. second twenty-five over-ground footfalls mat Kinematic heel-strike toe-off events used identify same cycles between systems. Nine by each system directly compared Measurements Bland-Altman methods, mean differences, Pearson correlation coefficients, intraclass coefficients. The results indicate measurements have good excellent agreement systems, except for stance time double limb support relative both systems stride width These findings adequately measure during