Validity and Test–retest Reliability of an Independent Self-Balanced, Socially Integrated Gait-Assisted Robotic System for Hip and Knee Joint Kinematics in Rehabilitation

Wearable robotic-assisted gait technologies have brought new possibilities for millions with gait dysfunction, yet a self-balancing robot that facilitates independent mobility is still lacking. We developed the Miracle, a walker-based self-balancing robot designed for community-dwelling ambulation, to fill this gap. However, the system’s concurrent validity and reliability had not been assessed prior to this study. This study aimed to establish the concurrent validity and reliability of the Miracle system’s kinematic measurements by comparing them with those from the gold standard Vicon gait analysis system. Our advanced Miracle robotic system, which incorporates artificial intelligence (AI)-enhanced sensing, was tested against the Vicon motion capture system, equipped with eight infrared cameras (Vicon Motion Systems, Oxford, UK), to validate measurements of hip and knee angular displacement. Data were analyzed using Python software. The Miracle system demonstrated excellent concurrent validity, with [Formula: see text] values ranging from 0.913 to 0.999, and strong test–retest reliability, with intraclass correlation coefficients between 0.927 and 0.982 for hip and knee joints under both slow and fast walking conditions. These findings confirm the Miracle system’s accuracy and reliability, supporting its clinical application in rehabilitation settings. The precise kinematic data it provides could enable clinicians to develop tailored gait training protocols and facilitate independent, safe ambulation for patients with gait impairments, thereby enhancing mobility and social integration in real-world environments.