Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2024 Oct 30;24(21):6998.
doi: 10.3390/s24216998.

VR-Aided Ankle Rehabilitation Decision-Making Based on Convolutional Gated Recurrent Neural Network

Hu Zhang  1 , Yujia Liao  1 , Chang Zhu  1 , Wei Meng  1 , Quan Liu  1 , Sheng Q Xie  2
Affiliations

VR-Aided Ankle Rehabilitation Decision-Making Based on Convolutional Gated Recurrent Neural Network

Hu Zhang et al. Sensors (Basel). .

Abstract

Traditional rehabilitation training for stroke patients with ankle joint issues typically relies on the expertise of physicians. However, when confronted with complex challenges, such as online decision-making or assessing rehabilitation progress, even seasoned experts may not anticipate all potential hurdles. A novel approach is necessary-one that effectively addresses these complexities without solely leaning on expert experience. Previous studies have introduced a rehabilitation assessment method based on fuzzy neural networks. This paper proposes a novel approach, which is a VR-aided ankle rehabilitation decision-making model based on a convolutional gated recurrent neural network. This model takes various inputs, including ankle dorsiflexion range of motion, angular velocity, jerk, and motion performance scores, gathered from wearable motion inertial sensors during virtual reality rehabilitation. To overcome the challenge of limited data, data augmentation techniques are employed. This allows for the simulation of five stages of rehabilitation based on the Brunnstrom staging scale, providing tailored control parameters for virtual training scenarios suited to patients at different stages of recovery. Experiments comparing the classification performance of convolutional neural networks and long short-term memory networks were conducted. The results were compelling: the optimized convolutional gated recurrent neural network outperformed both alternatives, boasting an average accuracy of 99.16% and a Macro-F1 score of 0.9786. Importantly, it demonstrated a strong correlation (correlation coefficient r > 0.9) with the assessments made by clinical rehabilitation experts, showing its effectiveness in real-world applications.

Keywords: convolutional gated recurrent neural network; rehabilitation; rehabilitation decision-making; stroke; whale optimization algorithm.

PubMed Disclaimer

Conflict of interest statement

The authors declare no conflicts of interest.

Figures

Figure 1
Figure 1
The decision-making algorithm implementation flowchart.
Figure 2
Figure 2
Virtual reality ankle rehabilitation scene. (a) Spacewalking. (b) Mountain hiking. (c) Flight simulation.
Figure 3
Figure 3
Flexible ankle rehabilitation robot platform.
Figure 4
Figure 4
Motion data collection based on Xsens DOT.
Figure 5
Figure 5
The different actual movements corresponding to virtual reality.
Figure 6
Figure 6
The effectiveness of collision detection for patients at various stages of rehabilitation.
Figure 7
Figure 7
The architecture of the WOA-CNN-GRU network.
Figure 8
Figure 8
Training loss and accuracy variations of deep learning models across different methods.
Figure 9
Figure 9
The accuracy of deep learning model.
Figure 10
Figure 10
Comparison of optimization algorithms using confusion matrices. (a) Training set (WOA). (b) Test set (WOA). (c) Training set (control group). (d) Test set (control group).
Figure 10
Figure 10
Comparison of optimization algorithms using confusion matrices. (a) Training set (WOA). (b) Test set (WOA). (c) Training set (control group). (d) Test set (control group).
Figure 11
Figure 11
The result of different optimization algorithms based on different test functions. (a) F1. (b) F2. (c) F6. (d) F7.
Figure 11
Figure 11
The result of different optimization algorithms based on different test functions. (a) F1. (b) F2. (c) F6. (d) F7.
See this image and copyright information in PMC

Similar articles

See all similar articles

References

    1. Marco G., Luciana L., Massimiliano M., Bragonzoni L., Knaflitz M., Benedetti M.G., Agostini V. Balance and Muscle Synergies during a Single-Limb Stance Task in Individuals with Chronic Ankle Instability. IEEE Trans. Neural Syst. Rehabil. Eng. 2023;31:4367–4375. - PubMed
    1. Ren Y., Wu Y.-N., Yang C.-Y., Xu T., Harvey R.L., Zhang L.-Q. Developing a Wearable Ankle Rehabilitation Robotic Device for in-Bed Acute Stroke Rehabilitation. IEEE Trans. Neural Syst. Rehabil. Eng. 2017;25:589–596. doi: 10.1109/TNSRE.2016.2584003. - DOI - PubMed
    1. Meghji M., Balloch A., Habibi D., Ahmad I., Hart N., Newton R., Weber J., Waqar A. An Algorithm for the Automatic Detection and Quantification of Athletes’ Change of Direction Incidents Using IMU Sensor Data. IEEE Sensors J. 2019;19:4518–4527. doi: 10.1109/JSEN.2019.2898449. - DOI
    1. Wang H., Lin M., Lin Z., Wang X., Niu J., Yu H., Zhang L., Vladareanu L. Virtual reality training system based on lower limb rehabilitation robot. Int. J. Eng. Technol. 2018;7:119–122. doi: 10.14419/ijet.v7i2.28.12892. - DOI
    1. Guzmán D.E., Rengifo C.F., Guzmán J.D., Cena C.E.G. Virtual reality games for cognitive rehabilitation of older adults: A review of adaptive games, domains and techniques. Virtual Real. 2024;28:1–17. doi: 10.1007/s10055-024-00968-3. - DOI

LinkOut - more resources