Survey on the research direction of EEG-based signal processing

Area of Science:

• Neuroscience
• Computer Science
• Biomedical Engineering

Background:

• Electroencephalography (EEG) is vital for Brain-Computer Interface (BCI) systems due to its portability and simplicity.
• Advancements in EEG signal processing are crucial for enhancing BCI performance.

Purpose of the Study:

• To comprehensively review EEG signal processing techniques (preprocessing, feature extraction, classification) since 2021.
• To identify trends, challenges, and future directions in EEG classification research.

Main Methods:

• Systematic review of 61 research articles from major academic databases.
• Analysis focused on preprocessing (including data augmentation), feature extraction, and classification methods.
• Emphasis on deep learning and multi-method fusion approaches.

Main Results:

• Preprocessing is widely adopted (96.7%) in EEG classification.
• Deep learning, particularly Convolutional Neural Networks (CNNs), shows significant promise.
• Multi-method fusion approaches are an emerging trend (49.2%).

Conclusions:

• Despite progress, challenges in EEG classification persist, including low cross-subject accuracy and interpretability.
• Further research is needed to address limitations in deep learning and fusion techniques.
• This review provides a foundation for improving EEG classification performance in BCI systems.