Arnaud Delorme

The 2025 EEG Foundation Challenge at NeurIPS is a brain-hacking quest where AI learns to decode thoughts across tasks and people, powered by datasets prepped with EEGLAB’s BIDS tools eeg2025.github.io youtube.com/watch?v=kSpnB_…

@DanFeuerriegel One of the paper's conclusions is that the pipelines do not differ much (which is a good thing). I agree that the reason for the difference between preprocessing pipelines should be further investigated -- especially when it is similar (e.g. different types of filtering).

@DanFeuerriegel There is no such thing as 'ground truth' dataset where the 'real' ERP differences are known.

@DanFeuerriegel Yes, this limitation is mentioned in the discussion.

New short lecture video about localization of ICA components #eeglab #ica youtube.com/watch?v=kQI3Yz…

A new EEGLAB video has been posted youtu.be/lCfYbwT0AA4

The second quarterly EEGLAB newsletter has been posted mailchi.mp/f005b076324e/t…

@robustgar the original paper may be found online. I am using machine learning on the images to assess potential bias before resubmitting it

@robustgar @deevybee control group irrelevant to null hypothesis. ES about 1.07> typical fMRI ES. Bayes requires priors. Rep would be great.

@robustgar @deevybee could you provide details about the statistical mistakes? The retraction mentions no methodological issue.