Computationally Assisted Upper Extremity Stroke Therapy Strategies Using Kinematic and Functional Measurements During Mixed Reality Rehabilitation
Margaret Duff, PhD
After stroke, the number and frequency of therapy sessions an individual receives may be determined by factors that are unrelated to the person’s specific therapy goals. This contributes, potentially, to the outcome that many people are being discharged with lasting disabilities. And during the therapy sessions that are received, the therapist must integrate the patient’s past and current ability with prior knowledge and expertise to best address the major impairments and realize functional goals in a short amount of time. Current research must focus on creating effective, data-driven therapy strategies that can both optimize time spent with the therapist in the clinic and allow the patient to continue therapy in a long-term, unsupervised setting, such as the home.
“Mixed Reality Rehabilitation”, a novel integration of repetitive reaching task training with high-resolution kinematic analysis and interactive media feedback, is well suited to begin to address the issues stated above. Mixed reality rehabilitation has previously been shown to generate measurable improvements in function and movement quality after stroke. However, the success of the therapy greatly relies on constant input from a therapist and the best schedule to accomplish long-term changes in motor strategy is still not well understood. Accordingly, the goal of this project is to create data-driven computational methods for administering mixed reality rehabilitation to achieve short-term learning, long-term retention and overall improvement in function and movement quality.