Biofeedback can increase motivation and improve movement quality during training, but motion capture systems rarely provide biomechanical data in real time. This study aimed to develop an open-source, real-time kinematic feedback system using Vicon motion capture that can also stream data to third-party video game platforms (e.g., Unity). The system used custom Python scripts to extract markers and device data from Nexus using the Plug-in Gait marker set and calculate real-time joint kinematics (i.e., sagittal/frontal hip, sagittal knee, and sagittal ankle angles). Twenty-eight healthy adults walked at 75%, 100%, and 125% of their self-selected walking speed. Agreement was evaluated against post-processed motion capture data using Statistical Parametric Mapping (SPM), intraclass correlation coefficients (ICC), and Bland-Altman plots. A pilot motor-learning experiment was conducted to evaluate how real-time feedback could be streamed to an open-source gaming platform. Our analyses showed that the real-time feedback system had errors when measuring sagittal hip (mean absolute angle MAE 0.99), knee (MAE 0.94), ankle (MAE 0.64), and frontal hip angles (MAE 0.89). During the pilot experiment, subjects utilized real-time feedback to alter their hip angle while walking. These results indicate that the system can provide low-latency estimates of selected gait kinematics during treadmill walking in healthy adults, with strongest agreement for hip and knee measures and more variable agreement for ankle angles. This open-source framework may support Vicon-based gait-feedback applications, but further validation is needed in clinical and rehabilitation settings.
Ackun et al. (Mon,) studied this question.