The purpose of the present study was to examine the repeatability of five algorithm-derived motion health screening scores (i.e., readiness, explosiveness, functionality, quality, and dysfunction) obtained from an innovative three-dimensional markerless motion capture system, composed of eight high-definition cameras recording at 60 fps. Thirteen females and six males performed two sets of three motion capture screenings, separated one week apart (six in total). The screenings consisted of 20 body movements performed in sequential order. Each screening within a testing session was separated by a 30 min rest interval to avoid the possible influence of fatigue. A trained research team member, facing the participant and standing outside of the camera capture range, was present to demonstrate each individual movement. The order in which motions were performed was identical across all participants. Repeated measures analysis of variance and intraclass correlation coefficients were used to examine statistically significant differences and measurement agreement across six testing sessions. The findings of the present study revealed no significant differences in algorithm-based motion health screening scores across multiple testing sessions. Moreover, excellent measurement reliability was found for readiness scores (ICC, 95% CI; 0.957, 0.914-0.980), good-to-excellent for functionality (0.905, 0.821-0.959) and explosiveness scores (0.906, 0.822-0.959), and moderate-to-excellent for dysfunction (0.829, 0.675-0.925) and quality scores (0.808, 0.635-0.915).
Keywords: assessment; biomechanics; exercise; human motion; measurement; technology.