FP-GCN: Frequency Pyramid Graph Convolutional Network for Enhancing Pathological Gait Classification
Xiaoheng Zhao1, Jia Li2, Chunsheng Hua1
1Institute of Intelligent Robots and Pattern Recognition, School of Cyber Science and Engineering, Liaoning University, Shenyang 110036, China.
Sensors (Basel, Switzerland)
|June 19, 2024
Summary
This study introduces a novel Frequency Pyramid Graph Convolutional Network (FP-GCN) for improved pathological gait classification. FP-GCN enhances temporal and spatial feature extraction, achieving high accuracy in identifying gait abnormalities.
Area of Science:
- Biomechanics
- Artificial Intelligence
- Medical Diagnostics
Background:
- Gait analysis provides crucial health insights but current methods struggle with temporal-spatial dependencies.
- Inefficiencies in pathological gait classification limit diagnostic accuracy.
Purpose of the Study:
- To propose a Frequency Pyramid Graph Convolutional Network (FP-GCN) for enhanced gait feature extraction.
- To improve the accuracy and efficiency of pathological gait classification.
Main Methods:
- Developed FP-GCN integrating spectral decomposition for temporal analysis and a pyramidal approach for spatial feature extraction.
- Utilized diverse public and proprietary datasets for model validation.
Main Results:
- FP-GCN achieved 98.78% accuracy on public datasets and 96.54% on proprietary data.
- The proposed method significantly surpasses existing methodologies in pathological gait recognition.
Conclusions:
- FP-GCN offers advanced feature extraction and recognition for pathological gait.
- This technology has potential applications in remote health monitoring and personalized interventions, especially in resource-limited settings.
