Brain Imaging Data Structure (BIDS) Steering Group · 2019–2022
The Brain Imaging Data Structure (BIDS) is a standard specifying the description of neuroimaging data in a filesystem hierarchy, and of the metadata associated with the imaging data. The PET BIDS extension (BEP0009) was released with BIDS v1.6.0.
What BIDS is for
The goal of BIDS is to make neuroimaging data more accessible, shareable, and usable by researchers. To achieve this, BIDS seeks a simple and intuitive way to organize and describe neuroimaging and associated data, guided by three foundational principles:
- Minimize complexity and facilitate adoption by reusing existing methods and technologies wherever possible.
- Tackle 80% of the most commonly used neuroimaging data, derivatives, and models (inspired by the Pareto principle).
- Adoption by the global neuroimaging community, and their input during the creation of the specification, is critical to the project's success.
How I got involved
In 2016, the NeuroReceptor Mapping conference (NRM2016) in Boston hosted a panel discussion on positron emission tomography (PET) data sharing. The audience unanimously supported a motion to establish a working group to propose guidelines for data sharing. Following this, in fall 2017 I attended NeuroHackWeek in Seattle and met Chris Gorgolewski. Chris had been called in by the PET community to help with data standardization and with adopting a PET BIDS extension – and when he heard I was interested in data sharing and worked with PET at one of the biggest PET labs worldwide, he asked me to help develop the PET BIDS extension (BEP0009).
In fall 2019, I became a member of the BIDS Steering Group. Membership is by election by BIDS contributors, and terms run for 2 years. I served until 2022, when I was succeeded by Cyril Pernet and Yaroslav Halchenko.
Why I ran for election
My biggest motivation for working on BIDS was facilitating data sharing. Compared to MRI and most other neuroimaging modalities, PET is expensive, technically challenging and invasive. This leads to small sample sizes (typically ~20–50 people) and to analysis workflows being highly dependent on the lab that acquired the data. I wanted to build momentum in the PET data-sharing world, so it was a natural fit to get involved in the steering group to represent PET.
My main focus
My main focus was outreach to communities not yet heavily involved in BIDS. Compared to the MRI community, the PET community needed to reinvent itself and embrace data sharing, as pointed out in a major PET journal. To that end, I was actively involved in organizing data-sharing/BIDS workshops (e.g. at NRM2020) and a BIDS derivative meeting in Copenhagen (2023), and worked to get more PET researchers interested in contributing to and using BIDS. As a steering group member, I also worked to identify further imaging modalities/derivatives in need of standardization and to support them on the road to BIDS.
This work continues today through the OpenNeuroPET project, and through my paper with Russell Poldrack, "Data sharing in neuroimaging: experiences from the BIDS project" (Nature Reviews Neuroscience, 2023) – see Publications.