TechRxiv
Main_file.pdf (934.52 kB)
Download file

Foot2hip: A deep neural network model for predicting lower limb kinematics from foot measurements

Download (934.52 kB)
preprint
posted on 21.03.2022, 04:06 by Rishabh BajpaiRishabh Bajpai, Deepak Joshi
Objective: This study aims to develop a neural network (foot2hip) for long-term recording of gait kinematics with improved user comfort. Methods: Foot2hip predicts ankle, knee, and hip joint angle profiles in the sagittal plane using foot kinematics and kinetics during walking. Foot2hip consists of three convolution, two max-pooling, two LSTM and three dense layers. An indigenously developed insole and an outsole were used to measure the kinetics and kinematics of the foot, respectively. Seven healthy participants were recruited to follow an experimental protocol consisting of six walking conditions: slow, medium, fast walking speed, rearfoot, flatfoot, and forefoot landing pattern. Results: When tested for leave-one-out and nested cross-validation, foot2hip obtained 3.04° ±0.20 RMSE and 0.97±0.01 correlation coefficient for knee joint, 1.7°± 0.09 RMSE and 0.95±0.01 correlation coefficient for hip joint, and 1.32°±0.08 RMSE and 0.91±0.02 correlation coefficient for ankle joint (averaged across all folds). Conclusion: The prediction performance of foot2hip is encouraging and shows its applicability in accurately predicting lower limb kinematics with minimal wearables. Significance: The hardware used along with foot2hip is low cost ($268 + N× $35, N is the number of different foot sizes), comfortable, and easy to use. Therefore, it is suitable for most clinical and personal applications

History

Email Address of Submitting Author

rishabhbajpai24@gmail.com

ORCID of Submitting Author

0000000205428255

Submitting Author's Institution

Indian Institute of Technology Delhi

Submitting Author's Country

India