Validation of Gait Kinematics With Ramp and Stair Ascent and Descent Revealed by Markerless Motion Capture in Simulated Living Space: Test-Retest Reliability Study.

Background: In recent years, there has been an increasing demand for markerless motion capture systems, which are being widely used in biomechanical and clinical research. Furthermore, by using a markerless motion capture system in a laboratory environment that mimics living spaces, the data acquired on various activities of daily living, such as level walking, ramp walking, and stair ascent and descent, should more closely resemble that of real-life activities. However, the absolute reliability of gait parameters in this context is still unclear. Objective: The aim of this study was to evaluate the reliability of a markerless motion capture system in assessing the ascent and descent of ramps and stairs during walking in a simulated living space. Methods: A total of 21 healthy participants performed level walking, ramp and stair ascent and descent on two separate days, with at least a 24-hour interval between sessions. Joint angles were measured using 27 synchronized cameras with a markerless motion capture application, Theia3D (Theia Markerless Inc), and analyzed in Visual3d for all planes of motion at the hip-, knee-, and ankle-joints. The absolute reliability of day-to-day reproducibility was assessed using full-curve analysis (root mean square difference [RMSD]) and discrete point analysis of gait events using the standard error of measurement (SEM). SEM was calculated only for level walking and ramp ascent and descent, where gait events were correctly detected. Results: The SEM values for level walking and ramp ascent and descent were all below the 5-degree threshold. However, while RMSD values were generally below 5°, this threshold was exceeded for knee-joint flexion-extension angles during ramp ascent and stair ascent (5.07° and 5.64°, respectively). Conclusions: The markerless motion capture system in the living laboratory setting demonstrated a high degree of accuracy for various environments and gait types. The low SEM values obtained indicate good reliability for joint angle measurements across different days. The slightly higher RMSD values for knee-joint angles during ramp and stair ascent may reflect the system's ability to capture the adaptations in joint kinematics in response to changes in gait conditions. These measurements in a living laboratory environment validated the absolute reliability of various gait parameters not only in level walking but also in ramp and stair ascent and descent. The findings suggest potential clinical applications and research opportunities, including the development of assistive devices and robots, using markerless motion capture in more natural living situations, rather than in controlled environments.