search

mne-tools / mne-bids

MNE-BIDS is a Python package that allows you to read and write BIDS-compatible datasets with the help of MNE-Python.
https://mne.tools/mne-bids/
BSD 3-Clause "New" or "Revised" License
132 stars 87 forks source link

Feature request: Support of derivatives #1176

Open christian-oreilly opened 1 year ago

christian-oreilly commented 1 year ago

Describe the problem

I am currently writing a grant proposal where I will suggest a aim that looks like this (the wording will probably change between now and the submission, but the idea should remain more or less the same):

Aim 1: Improve MNE-BIDS and EEGNET to support derivatives BIDS datasets of EEG/MEG source data. We will extend MNE to ease the creation and automated analyses of shared neurophysiological datasets. Particularly, we will focus on adding support for derivatives datasets, including EEG source data. We will demonstrate this integration in the context of EEGNET web platforms EEGNET and the PyLossless pipeline.

Describe your solution

For this grant, and for our PyLossless pipeline (https://github.com/lina-usc/pylossless), I would be interested in MNE-BIDS supporting more transparently BIDS derivatives, including the writing/loading of derivatives in subdirectories as outlined here: https://bids-specification.readthedocs.io/en/stable/03-modality-agnostic-files.html#derived-dataset-and-pipeline-description

I would also be interested in supporting more transparently BIDS mechanisms for storing EEG/MEG source data derivatives. I understand that this may also require, for consistency, discussion with the BIDS project to make sure that the norm itself supports these data, not just that software like MNE-BIDS supports an ad hoc version of it.

To be clear, this feature request ticket is not for one of the MNE-BIDS developers to write such a PR; the grant would support the salary for someone to contribute to this on my side. I would just want, at this step, to initiate a discussion so that I can know your level of interest in integrating such functionalities in MNE-BIDS and get your opinion about the best way to integrate these features.

Describe possible alternatives

None.

Additional context

No response

agramfort commented 1 year ago

awesome initiative !

so far the mne-bids-pipeline has hacked his way for derivatives and it's not done in a standardized way at all. I think it makes your project very relevant to move forward here.

as you said it's important to coordinate this with the wider EEG / MEG BIDS folks to make it adopted beyond mne-bids

Message ID: @.***>

hoechenberger commented 1 year ago

I'd like to second what Alex said: this is a great initiative! 👍👍👍

sappelhoff commented 1 year ago

sounds wonderful!