Hi everyone,
I am developing and validating a provenance-tracked preprocessing workflow for high-density EEG in MNE-Python. The workflow keeps automated QC prompts separate from manual scientific decisions. I have completed a conservative review of one public HBN recording, but three short intervals remain genuinely ambiguous after inspecting the static QC panels.
I am looking for an EEG researcher who would be willing to give a small, independent second opinion on these three intervals. The review should take about 15–20 minutes and would answer only three questions for each interval:
- Should the interval be kept or annotated as bad?
- If annotated, what are the approximate onset and offset in the zoomable raw signal?
- Should it be omitted from ICA fitting, from task epochs, or from both?
The unresolved prompts are:
- a broad transient or step-like change observed in E109 within a 100–120 s review window;
- a severe late artifact observed in E48 within a 240–253.176 s review window;
- overlapping late contamination observed in E49 within the same 240–253.176 s window.
The source dataset is public, but my derived review materials are currently controlled and unpublished. I would therefore first share only the minimum material required for the review and confirm an appropriate transfer method. I will preserve the independent decisions unchanged and report any disagreement transparently.
This is a voluntary, unpaid request with a deliberately small scope. If the review later becomes a substantive contribution to a paper or software validation, acknowledgment or authorship would be discussed according to the actual contribution rather than promised in advance.
If you have experience with manual EEG artifact annotation and would be willing to help, please reply here or send me a private message.
MNE-Python version: 1.12.1
Operating system: macOS