Quantitative assessment of lower limb and cane movement with wearable inertial sensors

Individuals with an age, injury, or disease-related mobility impairment often utilize a walking aid, such as a cane, to increase safety and stability during ambulation. Many individuals use a cane incorrectly and demonstrate altered gait patterns. Consequently, measuring the relationship between cane use and gait characteristics has potential to provide users, clinicians, and caregivers insightful information about cane-assisted walking. In this paper, we investigate fine-grained, objective measures of cane movement acquired from wearable inertial sensors. Specifically, we compute quantifications of swing and stance variability for both lower limbs and a cane device. We also introduce a novel visualization, the stance and swing phase plot, to facilitate insights into the sensor data. The computed gait parameters and visualization can potentially inform users and clinicians about assistive device usage over time and provide feedback about correct movement. We demonstrate the utility of the proposed algorithms with inertial sensor data collected from two patients undergoing inpatient stroke rehabilitation.