Related Experiment Video
Updated: May 13, 2026
06:07
Construction of Local Field Potential Microelectrodes for in vivo Recordings from Multiple Brain Structures Simultaneously
Published on: March 14, 2022
Dual-VCT: A dual-branch VMD-CNN-transformer model for local field potentials decoding
Xiao Li1, Yu Zeng1, Yongkang Zhou1
1Hubei Key Laboratory of Modern Manufacturing Quantity Engineering, School of Mechanical Engineering, Hubei University of Technology, Wuhan 430068, People's Republic of China.
Journal of Neural Engineering
|May 11, 2026
Summary
This study introduces Dual-VCT, a new model for brain-machine interface decoding that improves accuracy and robustness. Dual-VCT enhances local field potential (LFP) decoding for more stable and effective brain-computer interfaces.
Area of Science:
- Neuroscience and Biomedical Engineering
- Machine Learning for Neural Decoding
Background:
- Intracortical brain-machine interfaces (iBMIs) require robust Local Field Potential (LFP) decoding for clinical application.
- Existing decoding methods face limitations in feature utilization, multi-scale fusion, and robustness across different tasks and chronic conditions.
Purpose of the Study:
- To develop an advanced LFP decoding model addressing the limitations of current methods.
- To enhance the clinical translation and long-term stability of intracortical brain-machine interfaces.
Main Methods:
- Proposed Dual-VCT, a dual-branch Variational Mode Decomposition-Convolutional Neural Network-Transformer (VMD-CNN-Transformer) model for end-to-end LFP decoding.
- Implemented a symmetric time-frequency parallel architecture with independent VMD modules for temporal and frequency-domain signal decomposition.
- Utilized a hierarchical fusion pipeline for robust cross-scale feature integration.
Main Results:
- Achieved high classification accuracy (0.930±0.023) in a spatial grasping task and high correlation (0.910±0.023) in a finger tracking task in non-human primates.
- Significantly outperformed comparative dual-branch methods (p < 0.05) and showed a 4% performance gain over single-feature decoding.
- Demonstrated strong cross-task robustness and cross-day stability, with ablation studies confirming the efficacy of the dual-branch VMD design.
Conclusions:
- Dual-VCT offers a high-performance, structured paradigm for LFP decoding.
- The clinically oriented design supports the long-term stability crucial for chronic iBMI systems.
