How to participate

The Federated Journal Club is a trainee-oriented scientific reading activity transversal across the three Task Forces of WG3 (Automated Preprocessing Pipelines). Participants collaboratively read and critically annotate the neuroimaging literature across nine imaging modalities plus a cross-modality methods category, focusing on preprocessing.

New to reading papers critically? Start with our How to read a paper guide, then come back here to claim your first one.

Eligibility

The Journal Club is open to all interested practitioners, independently of their geographical location, experience, career stage or INDoS membership. We especially encourage trainees, early-career researchers and participants from Inclusiveness Target Countries. You do not need to be a registered INDoS member to take part. A free GitHub account is required to claim and track papers on the paper pool and leaderboard.

How claiming works

We have curated a pool of 242 preprocessing papers across nine imaging modalities plus a cross-modality methods category, all listed on the Board.

  1. Create a free GitHub account if you do not have one.
  2. Browse the pool and claim papers from that list (maximum three papers under review by a single participant at any given time). Papers with three reviews are pulled out of the pool, so reviews do not concentrate in a small subset.
  3. Get the paper. The Board provides URLs (to the DOI resolver). Should you not be able to access a paper, please contact us and we will send it to you.
  4. You have 12 days to read each paper you claim. Use a PDF annotation tool to leave inline comments. A larger comment at the end, wrapping up the most relevant aspects of the paper and your interpretation, is encouraged. Comments should mark what you did not understand, what has been contested or superseded since publication, and what matters for the INDoS mission of standardized, reproducible, shareable neuroimaging data.
  5. Upload the annotated PDF using that paper's upload link — you get one per paper in your claim thread, and it already knows which paper is yours.
  6. Sign it off. We check your file is readable and matches your claim, then say so on your claim thread and @-mention you. Reply /confirm <ID> and it goes to grading. That reply is your signature — see how to review a paper.
  7. Reviews are checked against a transparent rubric that drives a live leaderboard.

A few rules keep the pool flowing:

  • You may hold at most three papers at a time.
  • A paper stops accepting new claims once five people have claimed it, so that at least three completed reviews are expected.
  • A paper is complete once three people have reviewed it. The club stays open until every paper has three reviews.
  • If a paper is not for you, you can withdraw it at any time before the deadline. It simply goes back to the pool for someone else.
  • Partial reviews are not accepted. If you cannot finish, withdraw the paper rather than hand in an incomplete review.
  • Upload the annotated PDF before the deadline. An unmet deadline returns the paper to the pool automatically.
  • The clock stops when we receive your upload. From that point the paper cannot expire, so there is no rush on the /confirm and nothing at risk while you get to it.
  • A review that breaks these rules can be removed by the organisers, and the no-AI rule is the one that matters most. This applies after grading too.

How to review a paper

⛔ No AI — this is non-negotiable

Do not use AI tools to read, summarise, or annotate the papers. The entire point of the club is your own learning and judgement — the things AI cannot do for you.

Reviews produced with AI defeat that purpose, produce exactly the generic, low-criticality output the rubric scores lowest, and will not score. You declare, for every review you hand in, that the reading and the annotating are your own — by signing it off with /confirm <ID> on your claim thread, in your name.

Your review is the paper's PDF with inline annotations. Please follow these rules — they make reviews comparable, gradable, and usable for the final publication:

  • Annotate inside the PDF. Use the comment/annotation tool of your PDF reader (Acrobat, Preview, Okular, the Zotero reader, and similar) to attach comments to the exact passages they refer to.
  • Typed comments only. Annotations must be digital text comments. Scanned or photographed handwritten notes are not accepted — they cannot be read by the synthesis tooling or graded.
  • Anchor and spread your comments. Annotate across the whole paper (methods and results), not just the introduction, and point to the exact step or claim.
  • Cover the three angles. In your comments, mark (1) what you did not understand, (2) what has been contested, corrected, or superseded since publication, and (3) what is important for the Action (data sharing, standardization, reproducibility).
  • Sign your review off with /confirm <ID> once we tell you your upload has landed. The upload form asks you to affirm the no-AI rule as you go — but that form is open to anyone holding the link, so ticking it is a speed bump, not a signature. Your /confirm, posted from your GitHub account on your own claim thread, is what puts your name on the declaration. It is the one step organisers cannot take for you, and that is the point of it: it is what makes the club's one hard rule mean something. Nothing to remember, either — the bot @-mentions you, GitHub emails you, and you reply.

Your annotated PDF files are uploaded to our institutional survey platform at HES-SO Valais-Wallis (EU/Swiss data residency), visible only to the organisers. The materials later used in the final publication are the extracted, de-identified annotation dataset and methodology, never the copyrighted PDFs.

Each paper has its own upload link, posted in your claim thread when you claim it and repeated in every deadline reminder. The link carries the paper id with it, which is how your file is matched to your claim — so there is deliberately no generic upload link on this site, and a file sent any other way cannot be graded.

Grading and the leaderboard

Each completed review is scored by the organisers against a transparent rubric, with human spot-checks. The rubric rewards:

  • Engagement — comments spread across the whole paper.
  • Comprehension — specific, located "what I did not understand" notes.
  • Critical appraisal — contesting claims and flagging superseded or corrected work.
  • Relevance to the Action — the data-sharing, standardization, and reproducibility angle.
  • Originality — your own voice and judgement.

To grade fairly at the scale of hundreds of reviews and refresh the leaderboard daily, the organisers use AI assistance to score annotation quality against the rubric, calibrated by human checks. This is the one asymmetry in the club: organisers may use AI to grade; participants may not use AI to review.

The leaderboard refreshes daily and ranks participants by cumulative quality points, so finishing more good reviews moves you up. Withdrawals and incomplete reviews earn nothing. The standing in early August (date to be confirmed) is a major input to the Training School invitations.

Ready? Browse the pool and claim your papers →  ·  First, learn how to read a paper.