Design and Validation of a Brain-Controlled Hip Exoskeleton for Assisted Gait Rehabilitation Training.

This study presents an integrated micro-system solution to address the challenges of gait instability in patients with impaired hip motor function. We developed a novel wearable hip exoskeleton, where a flexible support unit and a parallel drive mechanism achieve self-alignment with the biological hip joint to minimize parasitic forces. The system is driven by an active brain-computer interface (BCI) that synergizes an augmented reality visual stimulation (AR-VS) paradigm for enhanced motor intent recognition with a high-performance decoding algorithm, all implemented on a real-time embedded processor. This integration of micro-sensors, control algorithms, and actuation enables the establishment of a gait phase-dependent hybrid controller that optimizes assistance. Online experiments demonstrated that the system assisted subjects in completing 10 gait cycles with an average task time of 37.94 s, a correlated instantaneous rate of 0.0428, and an effective output ratio of 82.17%. Compared to traditional models, the system achieved an 18.64% reduction in task time, a 28.31% decrease in instantaneous rate, and a 7.36% improvement in output ratio. This work demonstrates a significant advancement in intelligent micro-system platforms for human-centric rehabilitation robotics.