Age-sensitive differences in single and dual walking tasks from footprint floor sensor data

Abstract

Gait can provide insights of executive function decline. We present experiments and methodology for analysing age-sensitive differences in changes of walking patterns on 3 volunteers from three age groups: a young adult, an adult and a mature adult, by using an original footprint imaging floor sensor. The effect of cognitive load tasks in spatio-temporal walking patterns of the volunteers is captured in the experiments. Classification models based on Support Vector Machines (SVM) are applied to raw gait sensor data activities, including single tasks, such as normal and fast walk, as well as dual tasks. For single tasks, we report classifications with a top F-score of 93.36 ± 5.56. Competitive classification performance was obtained for the fine-grained walking variability in the dual task experiments.