Multimodal, Task-Aware Movement Assessment and Control: Clinic to the Home

The investigators seek to construct a novel, multimodal, distributed system that facilitates a new paradigm of home-based medical surveillance and treatment for frail older adults centered on timely diagnosis of movement system impairments and personalized intervention. Measurements from a heterogeneous set of complementary sensors will be combined with clinically-informed and data-learned dynamic models of human motion to enable real-time activity recognition (e.g., sitting, standing, walking) and movement assessments (e.g., speed, repetition, quality). As the study progresses, the system will be integrated with wearable assistive technology to provide "smart", activity-specific assistance of movement deficits or the activation of caregivers if a decline in function is detected. Finally, the investigators will test the combined assessment and intervention system in the clinic and the home, identifying challenges and solutions for the scaling up to unconstrained real-world settings.