3D ground reaction force (GRF) estimation during walking is important for gait and inverse dynamics analyses. Recent studies have estimated 3D GRF based on kinematics measured from optical or inertial motion capture systems without force plate measurement. A neural network (NN) could be used to estimate ground reaction forces. The NN network approach based on segment kinematics requires the selection of optimal inputs, including kinematics type and segments. This study aimed to select optimal input kinematics for implementing an NN for each foot’s GRF estimation. A two-stage NN consisting of a temporal convolution network for gait phase detection and a gated recurrent unit network was developed for GRF estimation. To implement the NN, we conducted level/inclined walking and level running on a force-sensing treadmill, collecting datasets from seven male participants across eight experimental conditions. Results of the input selection process indicated that the center of mass acceleration among six kinematics types and trunk, pelvis, thighs, and shanks among 15 individual segments showed the highest correlations with GRFs. Among four segment combinations, the combination of trunk, thighs, and shanks demonstrated the best performance (root mean squared errors: 0.28, 0.16, and 1.15 N/kg for anterior-posterior, medial-lateral, and vertical components, respectively).

In this study, an insole-type ground reaction force (GRF) measurement system using a load cell was manufactured and configured as a system that can measure joint angle and GRF, when walking in conjunction with a commercialized inertial sensor. The data acquisition device was used to acquire synchronized data, between the inertial measurement unit (IMU) sensor and the load cell insole. A three-dimensional motion analysis system comprising six infrared cameras and two ground reaction forces, was used to check the accuracy of the gait measurement system, comprising an inertial sensor and a load cell insole. The motion and force data were acquired while performing five times six-meter walking test by five young adult male subjects (Age: 26.0±1.8, Height: 171.4±6.8 cm, Weight: 62.2±10.8 kg). It was measured and as a result of comparing the calculated sagittal joint angle with the vertical GRF, the sagittal lower extremity joint angle correlation coefficient (Pearson’s r) was 0.40 to 0.94, and the vertical GRF to be 0.98 to 0.99. It is necessary to upgrade the joint angle calculation algorithm through future research. Additionally, the possibility of clinical application for actual stroke patients will be reviewed.