Regression transfer learning for the prediction of three-dimensional ground reaction forces and joint moments during gait
by Goksu Avdan; Sinan Onal; Banafsheh Rekabdar
International Journal of Biomedical Engineering and Technology (IJBET), Vol. 42, No. 4, 2023

Abstract: Clinical gait analysis is a useful tool for assessing a patient's walking conditions. Force platforms are gait analysis tools used to collect the ground reaction forces (GRFs); however, they are expensive and time-consuming. Therefore, this study focuses on the prediction of GRFs and joint moments without using force platforms. To address this problem, we proposed to combine deep learning methods with regression transfer learning (RTL). The inputs of the proposed method are joint angles and marker trajectories from a public dataset. Principal component analysis (PCA) has been used to reduce the data dimensionality to improve the computational time and prediction accuracy. A synthetic dataset has been generated to pre-train the deep learning method for transfer learning purpose. The experimental results indicate that the proposed transfer learning method increases the target domain's learning process and can successfully predict the average GRFs and joint moments with 97.44% and 96.56% accuracy, respectively.

Online publication date: Mon, 14-Aug-2023

The full text of this article is only available to individual subscribers or to users at subscribing institutions.

 
Existing subscribers:
Go to Inderscience Online Journals to access the Full Text of this article.

Pay per view:
If you are not a subscriber and you just want to read the full contents of this article, buy online access here.

Complimentary Subscribers, Editors or Members of the Editorial Board of the International Journal of Biomedical Engineering and Technology (IJBET):
Login with your Inderscience username and password:

Username:        Password:          Forgotten your password?

Want to subscribe?
A subscription gives you complete access to all articles in the current issue, as well as to all articles in the previous three years (where applicable). See our Orders page to subscribe.

If you still need assistance, please email subs@inderscience.com