Feature Selection, Construction, and Validation of a Lightweight Model for Foot Function Assessment During Gait With In-Shoe Motion Sensors

Identifying and monitoring overpronation oversupination in the long term during activities of daily living are essential for people’s ambulatory health. Using an in-shoe motion sensor (IMS) with power-saving functions is a potential solution. In this study, we challenged development estimation model foot function using center pressure excursion index (CPEI) as via linear multivariate regression, which sufficiently light type IMS. Data collected from 65 17 participants were involved construction validation, respectively. We validated ten scenarios simulating activities, including walking on different surfaces, shoes, or without carrying bag, indoors outdoors. applied statistical parametric mapping (SPM) to determine significant predictors performed our original feature selection algorithm, leave-one-subject-out least absolute shrinkage operator (LASSO), compress volume predictors. successfully discovered sex-specific constructed large effect-sized models that achieved high-precision CPEI estimation. estimated maximum 99.0% 100.0% males’ females’ data under same experimental conditions training 92.8%–100.0% 85.8%–100.0% scenarios. The effective possible provide applications long-term by