Project Description:

Gait patterns are valuable biomarkers for neurological conditions. However, instrumented gait analysis (motion capture, force plates) is infeasible outside the clinic. To address this, the Thakor Lab has developed an insole sensor with sufficient resolution and sampling rates for accurate assessment of everyday patient gait.

Our project involves the analysis of this insole sensor’s data for a healthy subject and a right leg paretic stroke patient across different walking conditions. We developed pressure-based and clinically-informed anatomical segmentation methods to identify foot regions and derive novel gait metrics. We assessed metric interpretability and ease-of-adoption based on literature review and clinician input. A one-way ANOVA test and error bars were used to evaluate left-right asymmetries, inter-subject variability, and rank each metric’s predictive value.

Overall, high-resolution, high-frequency foot pressure data reveals meaningful gait differences across and within individuals. These insights align with clinical expectations and may support personalized diagnosis and rehabilitation planning.

Project Photo:

A puffin writing code on a computer. Besides the keyboard is an insole, representing the pressure-sensing insole whose recorded data we seek to draw insights from.