Classifying sitting, standing, and walking using plantar force data

Abstract

Prolonged static weight-bearing at work may increase the risk of developing plantar fasciitis (PF). However, to establish a causal relationship between weight-bearing and PF, a low-cost objective measure of workplace behaviors is needed. This proof-of-concept study assesses the classification accuracy and sensitivity of low-resolution plantar pressure measurements in distinguishing workplace postures. Plantar pressure was measured using an in-shoe measurement system in eight healthy participants while sitting, standing, and walking. Data was resampled to simulate on/off characteristics of 24 plantar force sensitive resistors. The top 10 sensors were evaluated using leave-one-out cross-validation with machine learning algorithms: support vector machines (SVMs), decision tree (DT), discriminant analysis (DA), and k-nearest neighbors (KNN). SVM and DT best classified sitting, standing, and walking. High classification accuracy was obtained with five sensors (98.6% and 99.1% accuracy, respectively) and even a single sensor (98.4% and 98.4%, respectively). The central forefoot and the medial and lateral midfoot were the most important classification sensor locations. On/off plantar pressure measurements in the midfoot and central forefoot can accurately classify workplace postures. These results provide the foundation for a low-cost objective tool to classify and quantify sedentary workplace postures.

Keywords: Pattern recognition; Pedobarography; Plantar fasciitis; Posture differentiation; Weight-bearing.