Concurrent validity of Theia3D markerless motion capture for detecting sagittal kinematic differences between gait speeds

We do not yet understand the concurrent validity of markerless motion capture (MMC) to measure kinematic differences between multiple gait speeds. This study determined the capacity of Theia3D (Theia Markerless Inc.) MMC to detect sagittal-plane kinematic responses to different gait speeds during walking (3 and 5 km/h) and running (10 and 12 km/h). Fourteen participants ambulated on a motorised treadmill, while markerbased motion capture, through optoelectronic cameras (Oqus 7+, Qualisys AB), and MMC, through videos (Miqus, Qualisys AB) were synchronously collected. Sagittal plane changes in pelvis, hip, knee, and ankle kinematics were compared. Mostly excellent waveform similarity was found for joint kinematic changes (coefficient of
multiple determination [CMD] ≥ 0.87), but pelvic tilt was less similar (CMD ≤ 0.48). Agreement between outcome measures (joint minima and maxima, range of motion)
was mostly good-to-excellent (intraclass correlation coefficient [ICC] = 0.475-0.950) with standard error of measurement values of less than 1°. Pelvis kinematics showed lower agreement between systems (ICC = 0.032-0.776). In this study, Theia3D detected changes in hip, knee, and ankle sagittal-plane joint kinematics between speeds with a similar accuracy to the marker-based approach. Therefore, Theia3D is appropriate for use if interested in lower-limb sagittal joint kinematics, but not pelvic tilt.