Related Experiment Video
Updated: May 15, 2025
11:28
Concurrent EEG and Functional MRI Recording and Integration Analysis for Dynamic Cortical Activity Imaging
Published on: June 30, 2018
11.6K
Recognition of brain activities via graph-based long short-term memory-convolutional neural network
Yanling Yang1,2, Helong Zhao1,2, Zezhou Hao1,2
1School of Health Science and Engineering, University of Shanghai for Science and Technology, Shanghai, China.
Frontiers in Neuroscience
|April 8, 2025
Summary
A new graph-based long short-term memory-convolutional neural network (GLCNet) effectively classifies brain activities from magnetoencephalography (MEG) signals. This approach improves brain-computer interface (BCI) performance by handling individual variability in motor imagery (MI) and cognitive imagery (CI) tasks.
Area of Science:
- Neuroscience
- Biomedical Engineering
- Machine Learning
Background:
- Magnetoencephalography (MEG) offers high temporal and spatial resolution for neuroimaging.
- Classifying motor imagery (MI) and cognitive imagery (CI) from MEG signals is crucial for brain-computer interfaces (BCIs).
- Individual variability and signal disturbances pose significant challenges in brain activity recognition.
Purpose of the Study:
- To propose a novel deep learning model for robust classification of MI and CI tasks using MEG data.
- To enhance the accuracy and reliability of brain-computer interfaces (BCIs).
Main Methods:
- A graph-based long short-term memory-convolutional neural network (GLCNet) was developed.
- GLCNet integrates graph convolutional network (GCN), spatial convolution, and long short-term memory (LSTM) modules.
- The model extracts time-frequency-spatial features simultaneously for comprehensive analysis.
Main Results:
- GLCNet achieved superior performance compared to six benchmark algorithms on two public MEG datasets.
- Average accuracies of 78.65% (2-class) and 65.8% (4-class) were recorded on the MEG-BCI dataset.
- The model demonstrated robust performance in handling individual variability.
Conclusions:
- GLCNet significantly enhances the adaptability and robustness of brain activity classification.
- The findings contribute to advancing neuroscience research and brain-computer interface (BCI) applications.
