One of the most impenetrable problems challenging full recovery after stroke is the gap between motor capacity that a
stroke survivor regains (i.e. what they can do) and how they engage in home and community activities (i.e. what they
choose to do). To address this challenge, Flint Rehab developed MiGo, a novel multi-sensor activity tracker specifically
designed for stroke survivors. The unique feature of MiGo is the ability to capture and deliver feedback on both
quantitative and qualitative upper and lower limb activity of stroke survivors in their natural environment using the
same system. The long-term goal for this project and the MiGo technology is to develop a data-driven and clinically
informed behavioral intervention strategy that uses actionable quantitative and qualitative feedback to maximize
physical function after stroke.
This project combines Flint Rehab’s technology for real-time motion capture and expertise in developing
neurorehabilitation devices with the extensive experience in stroke neurorehabilitation of the Motor Behavior and
Neurorehabilitation team at the University of Southern California. This Phase I STTR aims to establish the feasibility,
validity, accuracy and usability of MiGo to monitor functional movement behaviors and deliver meaningful feedback to
stroke survivors across a broad range of motor impairments seen in this population (Aim 1). Specifically, 30 individuals
expressing a range (i.e., mild-severe) of motor impairment chronically after stroke will be recruited to participate in this
project. MiGo’s accuracy and utility cost to the end user will be assessed in a single in-lab session. Participants will be
outfitted with MiGo and perform functional standardized assessments. As gold-standard comparison, upper limb
movements will be compared to movement counts derived from a video recording of the standardized assessments,
whereas step counts and stance/stride time will be derived from the ADPM sensors during a walk test. Raw sensor data
from MiGo will be analyzed using proprietary algorithms for movement detection and compared to the gold-standards
to determine accuracy. Utility cost will be assessed using quantitative survey (social acceptability and ease-of-use),
sensor cost and time to don and doff each sensor to determine the minimal number of MiGo sensors and the optimal
placement on the body. In a subsequent step (Aim 2), short term feasibility and usability of the optimized MiGo will be
established by monitoring community-dwelling stroke survivors using MiGo over a 1-week interval in the natural
environment. Adherence, occurrence of adverse events, and satisfaction with MiGo will be recorded. Using the data
from the monitoring period, participants will be presented with a ‘Movement report’ with visual displays of quantitative
and qualitative feedback (Aim 3). Through survey and qualitative interview, we will identify the components of MiGo
feedback that users find most meaningful for driving lasting behavior change. The resulting technology will integrate
engineering and patient-centered rehabilitation approaches to promote fuller participation in meaningful life activities
outside clinical settings in a less structured environment—one where stroke survivors live their lives.
