Abstract

Movement disorders of the human foot-ankle complex are a common occurrence, owing to the altered joint mechanics during foot-ground interactions. Diagnostics of such movement disorders will require quantitative tools to evaluate in-vivo foot motions, in particular to the multi-segment/joint foot kinematics (MSFK), during gait. Unfortunately, current MSFK analysis largely rely on conventional technologies, such as skin-marker based motion capturing, video fluoroscopy and dynamic 3D scanning, being extremely time-consuming and costly. In this work, a novel movement tracking method, named the point-cloud foot analysis (PFA), was implemented with multi-view depth sensors, to allow fast evaluations of 3D motions of the foot-ankle complex during gait. Quantitative analysis obtained by the PFA methods and their accuracy relative to the conventional MSFK analysis methods were evaluated. The 3D surface reconstructions of the foot-ankle complex were achieved with a RMSE less than 2 mm. It was proven to be feasible to track multi-segment foot motions in both healthy and diseased subjects during walking conditions, with the processing time decreased from more than 4–6 h to less than 6 min for the entire flow of the contact phase analysis. The PFA method can be useful for fast evaluations of the movement disorders of the foot-ankle complex in diagnostics and design of therapeutic interventions and rehabilitation programs for clinical applications However, despite previous efforts, measurement accuracies of above methods may not be sufficient for the analysis of multi-segment foot motions, since the mean joint positioning errors of those methods were no less than 20 mm. To date, a feasible approach that allows efficient evaluation of 3D multi-segment/joint motions for the foot-ankle complex have not yet been established.