Prediction of Mobility After Stroke
Team: Fuchsia
- Program: Biomedical Engineering
- Course: Precision Care Medicine
Project Description:
Predictive models of mobility recovery and adverse event risk following stroke.
Project Photo:
Pink smart watch with a graph on the screen, with the title “Predicting Mobility Recovery After Stroke”
Project Poster
Project Post Summary:
Understanding patterns of mobility recovery after stroke remains an important problem for improving quality of life. This work focused on the use of minute-level Fitbit data (heart rate and step counts) to derive mobility features across the domains of heart rate, activity, and sedentary time. We then developed predictive models which demonstrate that these metrics can be used to predict performance on common clinic tests, such as comfortable and fast walking speed, as well as in the prediction of an individual’s risk for experiencing an adverse event. This work provides a baseline for the assessment of mobility outside of the clinic, and the potential for future development of an overall mobility score.