Towards better assessment and training of kinematics in post-stroke gait therapy




Shin, Sung Yul

Journal Title

Journal ISSN

Volume Title



Gait impairment is common following neurological injury such as stroke. Therapists train patients based on restoring healthy motions, or kinematics, but evidence for training proper kinematics is not well-established. Because the dosage of therapy has not been well quantified, it is unclear what aspects of gait therapy are important, but simply that more therapy is likely better. However, cost restrictions prevent such intensive therapy, incentivizing value-based care. Robotic gait trainers that repetitively train a specific kinematic walking motions can potentially ease the burden on therapists and allow greater patient throughput, improving the value of therapy. Still, the cost of these trainers is only affordable to the wealthiest clinics, leaving them unavailable to the vast majority of stroke survivors. The overall goal of my research is twofold: to show the role of kinematics in gait recovery following stroke, and how these kinematics can be trained in an economical manner. My first aim focuses on design of an affordable robotic gait trainer that can adapt to an individual’s healthy gait pattern. Such a device could make robotic gait training more accessible to new markets including resource-limited hospitals and even patients’ homes. My second aim presents development of an online algorithm for producing speed-dependent reference joint trajectories that can be used for general robotic gait training applications. The goals of my third and fourth aims investigate the importance of gait kinematics using a novel longitudinal cohort approach in subacute stroke patients. I quantified the dosage of therapy using a wearable motion capture to find correlates of functional recovery, defined as gait speed. I then questioned whether gait speed was sufficient to define gait recovery, taking an innovative look at how gait quality during this subacute period changes as gait function improves. I expect these aims will justify the importance of kinematics and suggest that wearable sensors can become a valuable tool for monitoring detailed kinematic motion, providing insight for more effective therapy regimens.


LCSH Subject Headings