Super-fast and accurate nonlinear foot deformation Prediction using graph neural networks

Taehyeon Kang; Jiho Kim; Hyobi Lee; Haeun Yum; Chani Kwon; Youngbin Lim; Sangryun Lee; Taeyong Lee

doi:10.1016/j.jmbbm.2024.106859

Super-fast and accurate nonlinear foot deformation Prediction using graph neural networks

J Mech Behav Biomed Mater. 2024 Dec 9:163:106859. doi: 10.1016/j.jmbbm.2024.106859. Online ahead of print.

Authors

Taehyeon Kang¹, Jiho Kim², Hyobi Lee², Haeun Yum³, Chani Kwon³, Youngbin Lim⁴, Sangryun Lee⁵, Taeyong Lee⁶

Affiliations

¹ Department of Mechanical and Biomedical Engineering, Ewha Womans University, Seoul, 03760, Republic of Korea; Department of Mechanical and Biomedical Engineering, Graduate Program in System Health Science and Engineering, Ewha Womans University, Seoul, 03760, Republic of Korea.
² Department of Mechanical and Biomedical Engineering, Ewha Womans University, Seoul, 03760, Republic of Korea.
³ Human Performance Lab Co., Ltd., #358, Research Cooperation Building, 52, Ewhayeodae-gil, Seodaemun-gu, Seoul, 03760, Republic of Korea.
⁴ SIMULIA, Dassault Systèmes Korea, Seoul, 06164, Republic of Korea.
⁵ Department of Mechanical and Biomedical Engineering, Ewha Womans University, Seoul, 03760, Republic of Korea; Department of Mechanical and Biomedical Engineering, Graduate Program in System Health Science and Engineering, Ewha Womans University, Seoul, 03760, Republic of Korea. Electronic address: [email protected].
⁶ Department of Mechanical and Biomedical Engineering, Ewha Womans University, Seoul, 03760, Republic of Korea; Department of Mechanical and Biomedical Engineering, Graduate Program in System Health Science and Engineering, Ewha Womans University, Seoul, 03760, Republic of Korea; Human Performance Lab Co., Ltd., #358, Research Cooperation Building, 52, Ewhayeodae-gil, Seodaemun-gu, Seoul, 03760, Republic of Korea. Electronic address: [email protected].

PMID: 39671974
DOI: 10.1016/j.jmbbm.2024.106859

Abstract

Recently, there has been a significant increase in the number of foot diseases, highlighting the importance of non-surgical treatments. Customized insoles, tailored to an individual's foot morphology, have emerged as a promising solution. However, the traditional design process of the customized insole is both slow and expensive due to the high computational complexity of finite element analysis (FEA) required to predict deformations of the foot. This study explores the applicability of a graph neural network (GNN) based on the MeshGraphNet framework to predict the 3-D shape of the foot under load and test the performance of GNN depending on the number of datasets. A total of 186 3-D undeformed foot CAD geometries are obtained from a series of 2-D foot images with deformations predicted through FEA. This FEA data is then used to train the GNN model, which aims to predict foot displacement with high accuracy and computation speed. After optimization of the weights of the GNN, the model remarkably outperformed FEA simulations in speed, being approximately 97.52 times faster, while maintaining high accuracy, with R² values above 95% in predicting foot displacement. This breakthrough suggests that GNN models can greatly improve the efficiency and reduce the cost of manufacturing customized insoles, providing a significant advancement in non-surgical treatment options for foot conditions.

Keywords: Computational modeling; Customized insole design; Foot deformation; Graph neural networks.