Evaluating a Wearable Sensor-Based Tibia Force Estimation Algorithm for Applications in Stress Fracture Reduction in Runners
Branscombe, Lauren M.
:
2018-05-16
Abstract
Tibia stress fractures are prevalent injuries in recreational runners, yet little is known about how to identify risk factors, due to a lack of ability to measure bone loading outside of a controlled lab environment. The goal of the present study is to determine if data from portable pressure insoles and an inertial measurement unit can act as surrogates for lab measurements and accurately and reliably estimate tibia bone loading using a modified inverse dynamics approach. Our hypothesis is that wearable estimates of tibia bone load can approximate lab estimates of tibia bone loading within a root mean square error of under ten percent peak loading for tibia force over stance, peak tibia force, and tibia load per kilometer. A three subject study was conducted to compare lab and wearable estimates of tibia force metrics, and results of the experiment support the hypothesis: with a simple calibration, wearable estimates were accurate within 6.1% for force across the entire stance phase, 3.1% for peak force, and 4.6% for load per kilometer. This level of accuracy, particularly for a preliminary study, is extremely promising, indicating that this method of calculating tibia bone loading could potentially be implemented in a portable system to record data from every day running and help to improve our fundamental understanding of tibia stress fracture injuries.