[home] [Personal Program] [Help]

---

Session: Poster session 2 (Odd numbers)
Session starts: Friday 27 January, 10:00
Presentation starts: 10:00

---

Bouke Scheltinga (University of Twente)
Jaap Buurke (Roessingh Research and Development)
Jasper Reenalda (University of Twente)

Abstract:
Running is a sport with a high injury incidence. Monitoring biomechanical load could help in understanding the development of these injuries. Ground reaction force (GRF) can be seen as an important measure to quantify biomechanical load during running [1]. However, GRF measurement is restricted to the lab oratory. Artificial neural networks (ANNs) are capable to model complex relations and thus could be used to predict GRF from inertial measurement units (IMUs) [2]. This prediction would be a step towards the quantification of biomechanical load outside the lab. The goal of this abstract is to show the possibilities of a generic ANN to estimate 3-dimensional GRFs using three IMUs. Methods: 11 experienced heel strike runners (4F, 7M; 30.6y ± 8.3y, 1.80m ± 0.11m, 74.2kg ± 17.4kg ) ran 9 trials on a force-instrumented treadmill at combination of three velocities (10, 12 and 14km/h) and three stride rates (preferred, -10% of preferred and +10% of preferred). Subjects were instrumented with IMUs (240Hz) mounted at both proximal tibias and pelvis. Using leave-one-subject-out cross validation, for each subject a model was created while data from the other subjects where used for training and validation of the model. Two layer ANNs (100 neuron s each) were then trained with the 3D gravity subtracted acceleration in the global frame as input to fit the 3D GRF. Performance of the models was analysed with the root mean squared error (RMSE) in body weight and as percentage of the range and Pearson’s correlation coefficient. Results: The ANN modelled GRFs with high accuracy for the forces in the sagittal plane with a RMSE ≤ 7% for the vertical direction and a RMSE ≤ 10% for the anterior-posterior direction for 8 out of the 11 subjects. Over all subjects, Pearson’s correlation coefficients of 0.96, 0.86 and 0.46 were achieved for respectively the vertical, anterior-posterior and medio-lateral direction. Conclusion: ANNs can be used to predict GRFs at a good accuracy for the vertical and anterior-posterior direction, but not yet in the medio-lateral direction. The variability in the performance for the different subjects could be explained by the variability in running gait between subjects. [1] Verheul, J., Nedergaard, N. J., Vanrenterghem, J., & Robinson, M. A. (2020). Measuring biomechanical loads in team sports–from lab to field. Science and Medicine in Football, 4(3), 246–252. https://doi.org/10.1080/24733938.2019.1709654 [2] Wouda, F. J., Giuberti, M., Bellusci, G., Maartens, E., Reenalda, J., van Beijnum, B.-J. F., & Veltink, P. H. (2018). Estimation of Vertical Ground Reaction Forces and Sagittal Knee Kinematics During Running Using Three Inertial Sensors. Frontiers in Physiology, 9, 1–14. https://doi.org/10.3389/fphys.2018.00218