Large EEG databases and automated processing solutions for scientific discovery and deep learning
Arnaud DELORME - UCSD (University of California San Diego, Swartz Center for Computational Neuroscience - SCCN) and Paul Sabatier of Toulouse University (CerCo et TMBI)
When: June 2nd
Where: Amphi Charve.
BIO : Arnaud Delorme’s primary research interest is in the analysis and modeling of high-dimensional EEG, MEG, and other imaging modalities. He is the originator of the widely used EEGLAB signal processing environment, the most used EEG signal processing software for 20 years in a row. He also played a leading role in the development of the NEMAR (Neuroelectromagnetic Data Archive and Tools Resource), which facilitates open sharing and standardized analysis of large-scale EEG and MEG datasets. His recent work includes systematic comparisons of automated EEG preprocessing pipelines, providing objective benchmarks for artifact rejection and data cleaning approaches, thereby advancing the reproducibility and scalability of neuroimaging analyses for deep learning applications.
Abstract: This study reviews publicly available EEG databases, focusing on NEMAR (Neuroelectromagnetic Data Archive and Tools Resource), and evaluates their accessibility, data quality, and potential for reuse. We examine the broader ecosystem of shared EEG resources, emphasizing the adoption of the BIDS (Brain Imaging Data Structure) framework as a standard for data organization and metadata annotation. Cloud-based processing solutions are assessed for their ability to support scalable and reproducible EEG analysis, particularly in conjunction with NEMAR. Additionally, we investigate EEGDash as a user-friendly interface that facilitates integration of EEG data with deep learning workflows, enabling efficient application of AI methods in neuroimaging research.
The programme:
This will be a NeuroSchool tutored seminar (the seminar will therefore be preceded and followed by discussions between PhD students in the PhD programme):
10:00 - Discussion among students to discuss the papers - For PhD students only
11:00 - Seminar - Open to all
12:00 - Special discussion time with the speaker for PhD students only
To take part in the discussions, PhD students must register on Ametice via this link : https://ametice.univ-amu.fr/mod/choice/view.php?id=4086340