Abstract
Accurate quantification of human motion is essential to understand mechanisms of performance, disease, or musculoskeletal injury. Historically, the field of human movement science has been limited by the availability of these tools, which has driven innovations in this field. The state of the art has vastly grown in the last half-century with many available systems for motion analysis. Consequently, all of these systems are limited to some extent, as most require tightly controlled laboratory environments, augmented scenes, considerable expertise or expense, and difficult numerical computations. The accuracy of a multi-view markerless motion analysis system was evaluated in this dissertation to determine its efficacy for use in clinical gait analysis. Compared to a gold-standard marker-based system, excellent validity was found in spatiotemporal parameters; but only limited correlations were seen between systems in kinematic variables. These results support the use of markerless-derived spatiotemporal parameters in healthy adults < 65, healthy adults > 65, and in those with Parkinson’s disease. Additionally, this work introduced a new approach for completely markerless inverse dynamics analysis, the results of which are promising for future work. Taken together, the future use of markerless technology for clinical gait analysis is feasible; but considerable work remains before this tool can be implemented in clinical practice.