Related Experiment Video
Updated: Dec 24, 2025
06:37
Electroencephalography Network Indices as Biomarkers of Upper Limb Impairment in Chronic Stroke
Published on: July 14, 2023
1.2K
EEG-based Classification of Lower Limb Motor Imagery with Brain Network Analysis
Lingyun Gu1, Zhenhua Yu2, Tian Ma2
1Key Laboratory of Child Development and Learning Science of Ministry of Education, School of Biological Science & Medical Engineering, Southeast University, Nanjing 210096, Jiangsu, PR China.
Neuroscience
|April 14, 2020
Summary
Researchers explored brain signal differences during left vs. right foot motor imagery. They identified distinct cortical network patterns, improving brain-computer interface (BCI) accuracy for lower limb control.
Area of Science:
- Neuroscience
- Biomedical Engineering
- Signal Processing
Background:
- Distinguishing between left and right foot motor imagery is crucial for advanced brain-computer interface (BCI) applications.
- Understanding the underlying cortical signal characteristics and network dynamics is essential for improving BCI performance.
Purpose of the Study:
- To investigate differences in cortical signal characteristics between left and right foot imaginary movements.
- To enhance the classification accuracy of motor imagery tasks using electroencephalography (EEG) data.
- To explore the network mechanisms underlying unilateral lower limb motor imagery discrimination.
Main Methods:
- Acquired 64-channel scalp electroencephalograms (EEGs) from 11 healthy participants during motor imagery tasks.
- Defined a cortical source model with 62 regions of interest over the sensorimotor cortex.
- Calculated functional connectivity using phase lock values for alpha (α) and beta (β) rhythms, applying network-based statistics and graph theory indices.
- Utilized sparse multinomial logistic regression (SMLR)-support vector machine (SVM) for feature selection and classification.
Main Results:
- Identified specific time-frequency differences in the alpha event-related desynchronization and beta event-related synchronization networks, congregating at the midline and involving premotor and primary somatosensory cortex.
- Observed statistically significant differences in network properties between left and right foot motor imagery tasks in alpha and beta rhythms.
- Achieved a maximum classification accuracy of 75% for single-trial discrimination between left and right foot imaginary movements using the SMLR-SVM model.
Conclusions:
- The study reveals distinct network mechanisms differentiating left and right foot motor imagery.
- These findings provide insights into the neural basis of unilateral lower limb motor imagery.
- The identified network characteristics offer a novel approach for improving BCI systems utilizing unilateral lower limb motor imagery.
