Related Experiment Video
Updated: Sep 13, 2025
09:57
Author Spotlight: Advancing Pediatric Epilepsy Surgery in Children Through Novel Biomarkers and Enhanced Localization
Published on: September 20, 2024
2.8K
Effect of EEG Electrode Numbers on Source Estimation in Motor Imagery
Mustafa Yazıcı1, Mustafa Ulutaş1, Mukadder Okuyan2
1Department of Computer Engineering, Faculty of Engineering, Karadeniz Technical University, Trabzon 61080, Türkiye.
Brain Sciences
|July 29, 2025
Summary
This study found that using 61 electroencephalogram (EEG) channels optimized motor imagery classification accuracy in brain-computer interface (BCI) applications. Fewer or more channels reduced classification performance, highlighting the importance of channel count.
Area of Science:
- Neuroscience
- Biomedical Engineering
- Signal Processing
Background:
- Electroencephalogram (EEG) is a key neurophysiological tool in neuroscience.
- EEG measurements vary in channel count for clinical and research use.
- Brain-computer interfaces (BCIs) leverage EEG for various applications.
Purpose of the Study:
- To investigate the impact of EEG channel count on motor imagery classification accuracy.
- To determine the optimal number of electrodes for BCI applications using source analysis.
- To evaluate classification performance across different channel configurations.
Main Methods:
- Utilized Common Spatial Patterns (CSP) as a spatiotemporal filter.
- Focused on mu band EEG signals sensitive to motor imagery.
- Tested channel configurations of 19, 30, 61, and 118 electrodes.
- Employed BCI Competition III Dataset Iva for experiments.
Main Results:
- The 19-channel configuration showed lower classification accuracy.
- 61 channels yielded the highest average classification accuracy (84.73%).
- 118 channels performed better than 19 but not as well as 30 or 61 channels.
Conclusions:
- The optimal number of EEG channels for motor imagery classification in BCI applications is approximately 61.
- Channel count significantly influences classification performance.
- Source analysis combined with CSP is effective for motor imagery BCI.
