1 of 25

Fully transparent study methodology descriptions

Remi Gau

Twitter: @RemiGau

eCOBIDAS

INCF assembly

​

12th September 2022

INCF eCOBIDAS working group

​

2 of 25

Origin story

3 of 25

Origin story

4 of 25

Origin story

5 of 25

COBIDAS

COmmittee on Best practice In Data Analysis & Sharing

Organization for Human Brain Mapping

6 of 25

COBIDAS

Organization for Human Brain Mapping

COmmittee on Best practice In Data Analysis & Sharing

7 of 25

  1. Experimental Design
  2. Acquisition​
  3. Preprocessing​
  4. Statistical Modeling & Inference ​
  5. Results​
  6. Data Sharing ​
  7. Reproducibility ​

​

COBIDAS report

40 pages long “checklist”

Mixing all the use-cases...

8 of 25

eCOBIDAS app

WEB

INTERFACE

MACHINE

FRIENDLY

FORMAT

COBIDAS

TABLE

9 of 25

eCOBIDAS app

REPROSCHEMA-UI

REPROSCHEMA

COBIDAS

TABLE

10 of 25

  • Cognitive and clinical questionnaires
  • Harmonize data by design
  • Repronim webinar

​

Reproschema

11 of 25

  • Unique and persistent identifiers
  • Versioning
  • Implementation agnostic
  • Uses a linked data graph (validated using SHACL)
  • Meant to work as a library ReproSchema Library

​

Reproschema

12 of 25

  • Unique and persistent identifiers
  • Versioning
  • Implementation agnostic
  • Uses a linked data graph (validated using SHACL)
  • Meant to work as a library ReproSchema Library

​

Reproschema

eCOBIDAS

Shared

reusable

items

13 of 25

User interface

(vue.js)

Reproschema

JSON-LD

Reproschema

Spreadsheet

Standardized

ouput

14 of 25

eCOBIDAS - neurovault

Neurovault collection metadata

  • 100 items
    • Subset of COBIDAS MRI
    • Reporting those → better than most papers

15 of 25

eCOBIDAS - neurovault

Neurovault collection metadata

DEMO

https://github.com/ohbm/cobidas

16 of 25

eCOBIDAS - eyetracker

  • Long term goal
    • Set of tools to facilitate the creation of similar checklists

cobidas-eyetracker

17 of 25

eCOBIDAS - data curation

ecobidas-PET

json

BIDS

  • Long term goal
    • Facilitate data curation

​

18 of 25

Future development

  • COBIDAS 2.0
    • Update of COBIDAS MRI coming in 2023
    • New and more diverse team
    • More modular
      • eg. refactor COBIDAS EEG/MRI
    • Checklist-oriented development

​

Peter

Bandettini

Shella

Keilholz

Janine

Bijsterbosch

Catie

Chang

Anqi

Qiu

Thomas

Yeo

Bertrand

Thirion

Jean-Baptiste

Poline

Chao Gan

Yan

Michael

Hanke

Thomas

Nichols

Cyril

Pernet

Henk

Mutsaerts

Samir

Das

Tristan

Glatard

T.J.H.

White

Remi

Gau

Jack

Van Horn

Russell

Poldrack

Athina

Tzovara

Ting

Xu

Alan

Evans

Dara

Cannon

Niko

Kriegeskorte

Simon

Eickhoff

19 of 25

Future development

  • Reproschema
    • Improve documentation
    • Improve tools to create new schemas

20 of 25

jsonld

Method section

templates

Future development

  • Long term goal
    • Automated method section generation
    • Import from existing standards

21 of 25

  • Difficulty: Easy
    • Improve response options (eg. add RRIDs)
  • Difficulty: Medium
    • Test one the apps to evaluate a paper
  • Difficulty: Hard
    • Help with UI development

How to help?

22 of 25

Thank you!

23 of 25

24 of 25

eCOBIDAS: aims

  • Provide human / machine readable descriptions of the data collection and analysis pipelines
  • Facilitate pre-registration and registered reports
  • Help make peer-review more systematic
  • Facilitate systematic literature reviews and meta-analyses
  • Facilitate data sharing

25 of 25

eCOBIDAS: full app 🐢

App

​

(("fMRI"[Title/Abstract] OR "functional MRI"[Title/Abstract] OR "functional magnetic resonance imaging"[Title/Abstract])

AND (humans[MeSH Terms]))

AND (("2017/01/01"[Date - Publication] : "3000"[Date - Publication]))

NOT ((Review[Publication Type]))

AND ((free full text[Filter]))

AND ((Neuron[Journal]))

AND ((english[Language])