EEGUnity: Open-Source Tool in Facilitating Unified EEG Datasets Toward Large-Scale EEG Model

Area of Science:

• Neuroscience
• Data Science
• Computational Biology

Background:

• The proliferation of Electroencephalogram (EEG) datasets and the rise of large-scale EEG models necessitate efficient data management tools.
• The heterogeneity of EEG data, including variations in content, metadata, and formats, presents significant challenges for data integration and research.

Purpose of the Study:

• To introduce EEGUnity, an open-source software tool engineered to address the complexities of managing diverse EEG datasets.
• To facilitate the efficient parsing, cleaning, and unification of multiple EEG datasets for improved data quality and consistency.

Main Methods:

• EEGUnity integrates four key modules: 'EEG Parser', 'Correction', 'Batch Processing', and 'Large Language Model Boost'.
• The tool employs intelligent data structure inference, data cleaning, and data unification techniques.
• It was evaluated using 25 diverse EEG datasets, showcasing various batch processing workflows.

Main Results:

• EEGUnity demonstrated high performance and flexibility in parsing and processing EEG data across multiple datasets.
• The tool successfully ensures data quality and consistency, crucial for reliable large-scale EEG research.
• The open-source nature and public availability of the code (github.com/Baizhige/EEGUnity) promote accessibility and collaboration.

Conclusions:

• EEGUnity provides a robust solution for managing the complexities of dispersed EEG datasets.
• The tool supports the advancement of large-scale EEG model research by offering a reliable foundation for data processing and integration.
• EEGUnity's modular design and demonstrated capabilities make it a valuable asset for neuroscientists and data scientists working with EEG data.