Summary:
“Stroke is the third leading cause of adult long-term motor disability. Gait rehabilitation approaching user-oriented and repetitive gait training has the potential for long-term mobility recovery. Active orthoses (AO) can tackle these rehabilitation abilities. More research is needed to foster gait rehabilitation oriented to the current user’s needs and to integrate wearable sensors into AOs for objective gait assessment. This thesis aims the development of SmartOs, a smart, modular, wearable active lower limb orthotic system, to foster user-oriented and repetitive gait training in impaired gait due to stroke and to evaluate human motor condition using kinematic and muscular gait measures. This work includes five research stages. First, a modular framework was implemented to integrate into an innovative and effective manner, wearable sensor systems, gait analysis tools, and control strategies into AOs. Second, a wearable motion lab including four wearable sensor systems, with an open-architecture for both stand-alone or third-party systems use, was successfully developed. The benchmarking analysis with commercial systems outlined that the sensor systems are purposeful for objective evaluation of the user’s motor condition. Third, a gait event detection tool through a finite state machine with an adaptive threshold-based structure was developed for detecting gait events in daily locomotion. Results show that the tool is suitable as a benchmark for detecting human gait events. Fourth, a machine learning-based recognition and prediction tool was achieved to classify locomotion modes and transitions. This tool advances the state-of-the-art by demonstrating that the exclusive use of kinematic data successfully classifies different walking directions and terrains. The last research stage made the SmartOs a multi-functional system through a hierarchical control architecture with four assistive control strategies. The user-oriented trajectory and adaptive impedance controls foster repetitive and assist-as-needed gait training, respectively. Both the EMG-based and user-orthosis interaction based control contribute to muscle strengthening. Findings indicate that SmartOs is functionally operative for a future clinical application as a personalized assistive and gait assessment solution of stroke survivors.”
Link:
