The purpose of this study is to develop a machine-learning-based regressor to estimate the gait-related parameters from the foot characteristics extracted by a foot scanning system. A fully-connected feed-forward neural network model was used to predict the gait parameters. The inputs of the model were the foot arch features and body anthropometric data, while the outputs of the model were the spatiotemporal gait parameters of the regular walking. The performance of the model was verified showing the accuracy of 95% or higher confirming the facts that foot features are dominant factors to estimate personalized gait patterns. In conclusion, the gait pattern can be easily assessed by measuring the foot depth-image from the foot scanner without using complex and expensive traditional methods if the data pools are significantly increased.
Mun, Kyung-Ryoul; Jeong, Hwansu; Oh, Hyungan; Hong, Junggi; Cho, Jae Yeong; and Kim, Jinwook
"A MACHINE-LEARNING-BASED GAIT ESTIMATION FROM THE FOOT ARCH PARAMETERS MEASURED BY A FOOT SCANNING SYSTEM,"
ISBS Proceedings Archive: Vol. 36:
1, Article 97.
Available at: https://commons.nmu.edu/isbs/vol36/iss1/97