Welcome | INDoS WG3 Training School 2026

name: title
layout: true
class: cover, indos-cover, no-logo

---
count: false

.cover-band[
<img class="lockup" src="../assets/indos-logo-left-white.png" alt="INDoS" />

.eyebrow[WG3 Training School · Madrid 2026]

# Welcome

### INDoS WG3 Training School on preprocessing pipelines

.who[Oscar Esteban <oscar.esteban@hes-so.ch> HES-SO Valais-Wallis · 30.09.2026]

<div class="mosaic"></div>
]

.cover-side[
<a href="https://www.indos-costaction.eu/wg3-ts2026-esteban/day1-01-welcome/">
 <object type="image/svg+xml" data="images/qr-talk-url.svg"></object>
 
 www.indos-costaction.eu/wg3-ts2026-esteban/day1-01-welcome/
</a>
]

.cover-logos[
<img src="../assets/cost-logo-darkgrey.png" alt="COST" /> <img src="../assets/eu-funded-pos.png" alt="Funded by the European Union" />
]

???

Welcome everyone, in the room and online, to the first Training School of INDoS Working Group 3.

We are hosted by CSIC here at Espacio Converge, Campus Serrano. The slides for this morning are online: scan the QR code, they are all under the same address.

---
name: newsection
layout: true

.perma-sidebar[

<a rel="license" href="http://creativecommons.org/licenses/by/4.0/"><img alt="Creative Commons License" style="border-width:0; height: 20px; padding-top: 6px;" src="https://i.creativecommons.org/l/by/4.0/88x31.png" /></a>
 INDoS WG3 Training School · Welcome (30.09.2026)

]

---

# This morning .gray-text[Wednesday 30 September]

.boxed-content.program-table[
| Time | Content |
|:--:|:--|
| 09:00 | .tag.plenary[Plenary] Welcome |
| 09:20 | .tag.plenary[Plenary] INDoS and Working Group 3 |
| 09:35 | .tag.plenary[Plenary] Introductions: 45 seconds each |
| 10:00 | .tag.theory[Theory] Standardized preprocessing: why pipelines, why automation |
| 10:45 | .tag.theory[Theory] BIDS across modalities |
| 11:15 | .gray-text[Coffee] |
| 11:30 | .tag.practice[Practice] BIDSManager hands-on, with Moustafa Almanla |
| 12:00 | .tag.theory[Theory] Containers |
| 12:30 | .tag.practice[Practice] Environment bring-up |
| 13:00 | .gray-text[Lunch] |
]

???

Block 1: standards, data, environments.

The afternoon (Block 2, with Mélanie Garcia) continues with Neurodesk (Steffen Bollmann, remotely from Toronto), how to read and review a neuroimaging paper, and the Journal Club.

Note the change at 11:30: the hands-on is BIDSManager, with Moustafa Almanla.

---

# Program

.cards.program.rows-2.cols-3.tight.compact[
.card.wed.accent[
.kicker[Wednesday · 09:00-13:00]

### Block 1 · Standards, data, environments

Standardized preprocessing, BIDS across modalities, BIDSManager hands-on, containers, environment bring-up.

.foot[Oscar Esteban, Moustafa Almanla]
]
.card.wed[
.kicker[Wednesday · 14:00-18:00]

### Block 2 · Environments, provenance, reading a paper

Neurodesk, how to read and review a neuroimaging paper, the Journal Club in practice.

.foot[Mélanie Garcia, with Steffen Bollmann (remote)]
]
.card.thu[
.kicker[Thursday · 09:00-13:00]

### Block 3 · PET, EEG/MEG and structural MRI

PET and PET-BIDS, EEG/MEG preprocessing and hands-on, FreeSurfer and VBM hands-on.

.foot[Martin Nørgaard (remote), Guiomar Niso, Yasser Alemán-Gómez]
]
.card.thu[
.kicker[Thursday · 14:00-18:00]

### Block 4 · Diffusion MRI

Diffusion MRI theory, then two hands-on sessions.

.foot[Yasser Alemán-Gómez]
]
.card.fri[
.kicker[Friday · 09:00-13:00]

### Block 5 · Functional MRI: did it work?

fMRI preprocessing with fMRIPrep, execution forensics, QA/QC hands-on.

.foot[Lea Waller]
]
.card.fri[
.kicker[Friday · 14:00-18:00]

### Block 6 · Personal projects and closing

Project pitches, project work on your own data, results, concluding remarks.

.foot[All trainers]
]
]

???

The full programme is on the school's page: www.indos-costaction.eu/training

Walk through the programme at a glance:

- Wednesday: standards, data and environments (morning); environments, provenance and how to read a paper, with the Journal Club (afternoon). Informal group dinner at 21:00, optional and at your own cost.
- Thursday: PET, EEG/MEG and structural MRI (morning); diffusion MRI (afternoon).
- Friday: functional MRI, did it work? (morning); personal projects and closing (afternoon).

Friday afternoon is for your own projects: bring your own data (BYOD).

---
class: vcenter

# Guest Wi-Fi

.boxed-content.center.large[
.eyebrow[Network]

.larger[**eventos-csic**]

.eyebrow[Password]

.larger[Given on site]
]

???

eduroam also works on the campus for those who have it set up.

---
class: vcenter

# Housekeeping

.cards.brand.cols-3[
.card[
.section-icon.corner[]

.kicker[Meals]

### Lunches and social dinner

**Lunch**: 13:00-14:00, every day.

**Group dinner**: tonight, Wednesday, at 21:00. Informal, optional, and at your own cost.
]
.card[
.section-icon.corner[]

.kicker[Every day]

### Attendance sheet

**Sign it every day** you attend.

Your signature is what entitles you to the daily allowance for that day: no signature, no reimbursement for the day.

.foot[All three days signed: the full 4.4 days (EUR 660).]
]
.card.alert[
.section-icon.corner[]

.kicker[After the school]

### Reimbursement

File your Online Travel Reimbursement Request (OTRR) in e-COST **by 14 October 2026, 23:59 CEST**.

Attach your tickets and receipts: once submitted, it cannot be modified.

.foot[COST allows 30 days, but we cannot guarantee processing after 14 October.]
]
]

???

Lunch: tell them where lunch is and how it works (to be completed on the day).

Dinner: where to meet for the 21:00 group dinner.

Attendance: in-person participants sign the attendance list every day. The daily allowance for Spain is EUR 150 a day, up to 4.4 days (100% for each school day attended, 100% for the travel-in day, 40% for the travel-home day), so up to EUR 660. The full amount needs all three days signed.

Reimbursement: the school's page asks for the OTRR within 12 days of the event, 14 October 2026 at 23:59 CEST. COST allows 30 days, but the Action cannot guarantee processing after 14 October. Travel is reimbursed at actual cost up to EUR 1,500, only for journeys over 100 km one way. Online attendance is not reimbursed.

---

# About us

.large[
**Electrophysiology** & **neuroimaging** researcher, **open science** advocate.
]

.about-photo.center[
<img src="images/guiomar.jpg" alt="Guiomar Niso" style="width: 82%; margin-top: 0.6em; filter: grayscale(1); border-radius: 10px;" />
]

.about-text.larger[
**Guiomar Niso** 
Head of the Neuroimaging Group 
Cajal Institute, CSIC, Madrid

* Vice-Chair of INDoS, co-leader of Working Group 3
* Member of the Young Academy of Spain (2022)
* Postdoc @ Indiana University and @ MNI (McGill)
* Ph.D. (2013) @ Universidad Politécnica de Madrid
* First author of MEG-BIDS
]

???

Guiomar Niso hosts the school here at CSIC. She leads the Neuroimaging Group at the Cajal Institute, is Vice-Chair of INDoS and co-leads Working Group 3.

Her work spans human electrophysiology and multimodal neuroimaging, and open science: the Open MEG Archives (OMEGA), Brainstorm and brainlife, and the BIDS standard (first author of MEG-BIDS).

She gives the EEG/MEG session on Thursday, online.

---

# About us

.large[
**Data scientist** who plays with brain data & **open science** advocate.
]

.about-photo.center[
<img src="https://www.axonlab.org/images/teampic/Oscar_HQD-1-square.jpg" alt="Oscar Esteban" style="width: 82%; margin-top: 0.6em; filter: grayscale(1); border-radius: 10px;" />
]

.about-text.larger[
**Oscar Esteban** 
Associate Professor & Head of [AxonLab](https://www.axonlab.org) 
School of Engineering, HES-SO Valais/Wallis

* Leader of INDoS Working Group 3
* Research & Teaching FNS Fellow (2025) @ CHUV
* PD (2020) @ Stanford University
* Ph.D. (2015) @ Universidad Politécnica de Madrid
* ESKAS (2012) @ EPFL
]

???

Quick context: I lead *fMRIPrep* and *MRIQC* and coordinate the NiPreps community. Within INDoS, I lead Working Group 3, Automated Preprocessing Pipelines, which organizes this school.

---

# About us

.large[
The **trainers** and the **assistant** of this school.
]

.cards.team.brand.cols-3.tight.compact[
.card[
.team-head[<img class="team-photo" src="images/team-steffen.jpg" alt="Steffen Bollmann" /> Steffen Bollmann]

Neurodesk, University of Queensland, Australia

.foot[Wed 14:00 · Neurodesk (remote)]
]
.card[
.team-head[<img class="team-photo" src="images/team-avatar.jpg" alt="Mélanie Garcia" /> Mélanie Garcia]

Massachusetts General Hospital, Harvard Medical School, USA

.foot[Wed afternoon · Reading papers, Journal Club]
]
.card[
.team-head[<img class="team-photo" src="images/team-avatar.jpg" alt="Martin Nørgaard" /> Martin Nørgaard]

University of Copenhagen, Denmark

.foot[Thu 09:00 · PET and PET-BIDS]
]
.card[
.team-head[<img class="team-photo" src="images/team-avatar.jpg" alt="Yasser Alemán-Gómez" /> Yasser Alemán-Gómez]

Lausanne University Hospital and University of Lausanne, Switzerland

.foot[Thursday · Structural and diffusion MRI]
]
.card[
.team-head[<img class="team-photo" src="images/team-lea.jpg" alt="Lea Waller" /> Lea Waller]

CRIUGM and Université de Montréal, Canada

.foot[Fri morning · fMRIPrep, QA/QC]
]
.card[
.team-head[<img class="team-photo" src="images/team-avatar.jpg" alt="Daniel Martín Díez" /> Daniel Martín Díez]

Complutense University of Madrid, Spain

.foot[Assistant]
]
]

???

The rest of the team, in programme order:

- Steffen Bollmann (Neurodesk, University of Queensland) opens this afternoon with Neurodesk, remotely from Toronto.
- Mélanie Garcia (MGH, Harvard Medical School) leads this afternoon: how to read and review a neuroimaging paper, and the Journal Club.
- Martin Nørgaard (University of Copenhagen) opens Thursday with PET and PET-BIDS.
- Yasser Alemán-Gómez (CHUV and University of Lausanne) takes most of Thursday: structural MRI (FreeSurfer, VBM) and diffusion MRI.
- Lea Waller (CRIUGM and Université de Montréal) leads Friday morning: fMRI preprocessing with fMRIPrep, execution forensics, QA/QC.
- Daniel Martín Díez (Complutense University of Madrid) is the school's assistant.

---
class: vcenter

# INDoS at a glance

.cards.brand.cols-3[
.card[
.kicker[COST Action CA24161]

### Improving Neuroimaging Data for Sharing

Make neuroimaging data truly FAIR (Findable, Accessible, Interoperable, Reusable) by developing shared standards, tools, and ethical guidelines for data sharing.

.foot[30 September 2025 to 29 September 2029]
]
.card[
.kicker[Network]

.metric[243]

.center[members from **38 countries**]

.foot[Listed on the Action's website, September 2026]
]
.card[
.kicker[Leadership]

### Chair

Jochem Rieger .gray-text[(University of Oldenburg)]

### Vice-Chair

Guiomar Niso .gray-text[(CSIC, our host)]

.foot[www.indos-costaction.eu]
]
]

???

COST Actions are networks funded by COST (European Cooperation in Science and Technology): they fund meetings, training schools, short-term scientific missions and dissemination, not research itself.

INDoS runs four years, from September 2025 to September 2029.

---
class: vcenter

# Four challenges

.boxed-content[
.large[Researchers who share, or use shared, neuroimaging data face four interrelated challenges:]
]

.cards.brand.cols-4[
.card[
.section-icon.corner[]

.kicker[WG1]

### Data quality

No standard for assessing data quality.
]
.card[
.section-icon.corner[]

.kicker[WG2]

### Metadata

Metadata are missing.
]
.card[
.section-icon.corner[]

.kicker[WG3]

### Preprocessing

No standard preprocessing pipelines.
]
.card[
.section-icon.corner[]

.kicker[WG4]

### Ethics and law

Unclear ethics and legal regulations of data sharing.
]
]

???

This is the Action's challenge, as written in its Memorandum of Understanding: "lacking standardization of data quality assessment and preprocessing pipelines, missing metadata, and unclear data sharing ethics and legal regulations."

Each challenge has a working group. WG5 takes care of dissemination and communication across all of them.

---
class: vcenter

# Five working groups

.cards.brand.cols-5.tight.compact[
.card[
.kicker[WG1]

### Quality Assessment and Quality Control

Recommendations, standard measures and reporting guidelines for QA/QC across MEG, EEG, MRI and fNIRS.

.foot[Lead: Robert Oostenveld]
]
.card[
.kicker[WG2]

### Metadata

Standardized metadata: extend BIDS, describe experiments, integrate metadata in tools, train people to use it.

.foot[Lead: Lyuba Zehl]
]
.card.accent[
.kicker[WG3]

### Automated Preprocessing Pipelines

Guidelines and tools for reproducible, documented, well-justified preprocessing. MEG and EEG first.

.foot[Lead: Oscar Esteban, co-lead: Guiomar Niso]
]
.card[
.kicker[WG4]

### Ethics, Practices, and Regulations

Tools and practices that lift the legal uncertainty around sharing human neuroimaging data.

.foot[Lead: Camille Maumet]
]
.card[
.kicker[WG5]

### Dissemination and Communication

Coordinates the Action's dissemination and communication activities.

.foot[Lead: Mathew Abrams]
]
]

???

The aims are paraphrased from the Memorandum of Understanding.

Co-leads: WG1 Michal Mikl and Karel Lopez Vilaret; WG2 Karel Lopez Vilaret, Anđela Šoškić, Başak Esin Köktürk Güzel and Mateusz Pawlik; WG4 Jochem Rieger; WG5 Esin Ozturk-Isik.

Anyone can join a working group: ask us how.

---

# Working groups organize in task forces

.cards.brand.cols-3[
.card[
.kicker[WG1 · QA/QC]

- Existing practices in neuroimaging
- Practices in other domains
- EEG quality assessment
- MEG quality assessment
- MRI quality assessment
]
.card[
.kicker[WG2 · Metadata]

- BIDS outreach and augmentation
- Guidelines on how to best use BIDS
- Tool alignment
- Demonstrator datasets
- Training materials
]
.card[
.kicker[WG4 · Ethics and regulations]

- Anonymity and reidentification
- Ethics, open science and data sharing
- Clinical and non-clinical aspects
- Legal aspects
]
]

.boxed-content[
.gray-text[Task forces are where the work happens: small teams, one concrete objective, open to any member.]
]

???

WG5 works across the Action and has no task forces of its own.

The WG2 tool-alignment task force focuses on tools that create or work with BIDS-compliant datasets, which is very close to what we do this morning.

---

# WG3: Automated Preprocessing Pipelines

.boxed-content[
.cta[
**Aim**: guidelines and tools for reproducible, well-documented, and well-justified preprocessing of neuroimaging data. The initial focus is on **MEG and EEG**, drawing on the more developed standardization of MRI preprocessing.
]
]

.cards.steps.tight[
.card[
.enum[T1]

**Inventory** of modality-specific preprocessing best practices, tools and pipelines (MRI, MEG, EEG, fNIRS).
]
.card[
.enum[T2]

**Networking** towards standards for rational preprocessing steps, with MEG/EEG in focus, reaching out to fMRIPrep and NiPreps.
]
.card[
.enum[T3]

**Documentation standards** for preprocessing and provenance tracking (with COBIDAS, NIDM, BIDS-Prov).
]
.card[
.enum[T4]

**Pipeline integration** opportunities (FieldTrip, MNE, NiPreps, fMRIPrep).
]
.card[
.enum[T5]

**Training material** for rational preprocessing. .gray-text[You are here.]
]
]

???

These are the five tasks written into the Memorandum of Understanding.

---
class: vcenter

# WG3 task forces and deliverables

.cards.brand.cols-3[
.card[
.section-icon.corner[]

.kicker[TF1]

### Inventory of tools and workflows

A taxonomy and survey of preprocessing tools and workflows across modalities.

.foot[Lead: Friedrich Carrle · D3.1: an up-to-date inventory, published on the Action's website]
]
.card[
.section-icon.corner[]

.kicker[TF2]

### Literature review

A pre-registered systematic review of preprocessing practices and how they are reported.

.foot[D3.2: open publications on rational preprocessing and provenance tracking]
]
.card[
.section-icon.corner[]

.kicker[TF3]

### Training materials

Preprocessing training resources for the community, **including this school**.

.foot[Lead: Başak Esin Köktürk Güzel · D3.3: training material, published on the Action's website]
]
]

.boxed-content[
.center.gray-text[Transversal to the three task forces: the **Federated Journal Club**.]
]

???

The deliverables D3.1 to D3.3 are those of the Memorandum of Understanding.

Everything you produce here (slides, notebooks, your own project on Friday) feeds D3.3.

---

# The Federated Journal Club

.left-column2[
.cards.cols-1.tight[
.card[
.metric[239]

.center[papers in the pool]
]
.card[
.metric[10]

.center[nine modalities + cross-modality methods]
]
]
]

.right-column2[
Participants collaboratively **read and critically annotate** the neuroimaging preprocessing literature.

- **Open to everyone**: no INDoS membership required.
- Especially welcomes **trainees and early-career researchers**.
- Anatomical MRI, EEG, CT, PET, MEG, fNIRS, fMRI, diffusion MRI, SPECT, and cross-modality methods.
- The annotations become an **open meta-review** publication, with participating reviewers credited as **co-authors**.

.center[
<a href="https://www.indos-costaction.eu/journal-club/">
<object type="image/svg+xml" data="images/qr-journal-club.svg" style="width: 22%"></object>
 
www.indos-costaction.eu/journal-club
</a>
]
]

???

The pool counts come from the Journal Club's live status on 30 September 2026.

The Journal Club is transversal to WG3's task forces: it feeds the literature review (TF2) and trains reviewers (TF3).

---

# How the Journal Club works

.cards.steps[
.card[
.enum[1]

**Claim** up to 3 papers at a time. Each claim runs for 12 days, with one 7-day extension.
]
.card[
.enum[2]

**Annotate** the PDF inline: what you did not understand, what has been contested or superseded, what matters for INDoS (data sharing, standardization, reproducibility).
]
.card[
.enum[3]

**Sign it off** from your GitHub account. **No AI** may be used to read, summarize, or annotate.
]
.card[
.enum[4]

**Graded** on a 5-axis rubric, with critical appraisal weighing most. A paper is done at 3 reviews.
]
]

.boxed-content[
.cta[
**This afternoon**: 15:00 How to read and review a neuroimaging paper, with Mélanie Garcia · 16:00-17:45 Journal Club, in practice.
]
]

???

Rules as enforced by the tracking system (RULES.md in the journal-club repository): at most 3 active claims per person, a paper closes to new claims once 5 people hold it, done at 3 completed reviews, quality floor of 2.0 out of 5.

The five axes: engagement and coverage, comprehension, critical appraisal, action relevance, originality (non-AI).

---
class: roulette
time: 45
include_org: true

# Introductions

.boxed-content[
.pull-left.large[
**45 seconds** each:

.no-bullet[
* Your **name**
* Your **affiliation** and country
* Your **career stage**
* Your **areas of interest**
* Your **motivation** to be here, and what you **expect** from the school
]
]

.pull-right[

.cta[
 **Joining online?** Introduce yourself in the chat with the same five points while the room goes around.
]
]
]

???

Space starts or pauses the roulette, S skips a name, K ends it. The time per person can be changed in the controls bar.

In-person introductions only: 28 people at 45 seconds is about 21 minutes of a 25-minute slot.

The names come from roster.local.yml, which is not in the repository: serve the deck locally for the real list.

---
class: vcenter

# Up next

.boxed-content[
.cta[
**10:00 Standardized preprocessing**: why pipelines, why automation, the pipeline end to end with fMRIPrep as the reference model.
]
]

.boxed-content.center[
All materials of this morning: **[www.indos-costaction.eu/wg3-ts2026-esteban](https://www.indos-costaction.eu/wg3-ts2026-esteban/)**, CC BY 4.0.

.large[

[](#1)
[](../day1-02-preprocessing/index.html)
]
]

.cost-ack[
<img src="../assets/cost-logo-darkgrey.png" alt="COST" /> <img src="../assets/eu-funded-pos.png" alt="Funded by the European Union" /> This publication is based upon work from COST Action CA24161 (INDoS), supported by COST (European Cooperation in Science and Technology).
]