A Functional Upper Limb Training and Assessment Tool to Enhance Efficacy and Scalability of Rehabilitation in Ecological Environments

ABSTRACT Impairments following stroke make it one of the leading causes of disability. Many individuals with stroke do not recover complete function of the upper limb at time of discharge from clinical services . Moreover, early stage improvements may wane following the cessation of formal therapies. Regaining as much upper limb function as possible is important, as even mild impairments are associated with limitations in daily function and lower health- related quality of life. Thus, finding ways to augment functional mobility is important. The overarching purpose of this project is to use portable technology, affordable for home use, to provide objective feedback on performance of upper limb motor tasks to individuals with residual deficits following chronic conditions such as stroke. Objective feedback serves to better inform the participant of their progress and actively engage them in their rehabilitation, thus encouraging self-management of rehabilitation. Results from a recent survey shows therapists predominantly provide patients with stroke written home exercise programs at time of discharge from therapies. With this static approach, patients have a limited capacity to evaluate their motor performance and no encouragement to refine their movement. Creating automated systems that measure and provide feedback will promote a more proactive approach to maximizing function across the lifespan of individuals with chronic conditions. We propose that coupling smart technology found in readily available smartphones with three-dimensional (3D) printing provides a scalable option to provide feedback in long-term rehabilitation and advances the current standard of care, i.e. written home exercise programs. We will build upon our current research, further developing and validating a novel in-home rehabilitation system using 3D printing technologies, smart devices and machine learning algorithms. Specifically, we will use emerging low- cost 3D printing techniques to augment off-the-shelf smart devices (e.g., smartphones) into functional rehabilitation tools. The built-in sensors and our machine learning apps in smartphones can quantify characteristics of movement and provide actionable feedbacks to users during in-home rehabilitation. We hypothesize that feedback will result in improved functional mobility. We will examine performance changes in common activities of daily living (ADLs) trained with our system. With the portable devices, we are able to go further than reporting the typical time metric and also provide feedback on smoothness of movement represented as a normalized cumulative jerk (rate of change of acceleration) score. This assessment provides information on control of movement, an indication of impairment. Recording normalized jerk over time and across activities is a promising way to assess change in impairment. Overall, this scalable design offers an innovative approach to long-term rehabilitation needs of individuals with stroke.