From 7fcd842bd9af40852bb25546a9be8f9319fd8170 Mon Sep 17 00:00:00 2001 From: Stefan Appelhoff Date: Fri, 12 Jan 2024 11:23:57 +0100 Subject: [PATCH] carry over initial text --- mkdocs.yml | 1 + src/derivatives/ephys.md | 43 ++++++++++++++++++++++++++++++++++++++++ 2 files changed, 44 insertions(+) create mode 100644 src/derivatives/ephys.md diff --git a/mkdocs.yml b/mkdocs.yml index 82badc4288..ecfeac3a9b 100644 --- a/mkdocs.yml +++ b/mkdocs.yml @@ -23,6 +23,7 @@ nav: - BIDS Derivatives: derivatives/introduction.md - Common data types and metadata: derivatives/common-data-types.md - Imaging data types: derivatives/imaging.md + - Electrophysiology data types: derivatives/ephys.md - Longitudinal and multi-site studies: longitudinal-and-multi-site-studies.md - Glossary: glossary.md - BIDS Extension Proposals: extensions.md diff --git a/src/derivatives/ephys.md b/src/derivatives/ephys.md new file mode 100644 index 0000000000..f1905c00a3 --- /dev/null +++ b/src/derivatives/ephys.md @@ -0,0 +1,43 @@ +# Electrophysiology data types + +This section pertains to human electrophysiological data, including EEG, MEG and iEEG, which characteristically have channels and time. +This includes minimally processed data, but excludes data that has undergone extensive processing, such as source reconstruction. + +## Minimally processed electrophysiological data + +A minimally processed EEG, MEG or iEEG data can be stored as a derivative, using the same file formats as specified for raw EEG, MEG, or iEEG. +Examples for this are MaxFiltered MEG data, or re-referenced and bandpass-filtered EEG data. +Certain file formats that are allowed in BIDS for raw data have limited representation to some extent, +for example the EDF and BDF formats cannot be used to store epoched (or "segmented") data in a standardized way, +whereas the BrainVision and FIF file formats can. +These limitations should be taken into account when writing derivative data back to disk. +The representation of the data MUST follow the general derivative conventions and the Common file level metadata fields, +notably the use of the [`desc` entity](../appendices/entities.md#desc). + +If minimally processed data is the result from processing a BIDS dataset, then it MUST be marked as a derivative and the source data must be specified. +The processing steps MUST be indicated in the [`desc` entity](../appendices/entities.md#desc), +where the label is a description identifier that distinguishes it from the original raw data. + +If minimal processing was performed on the source data, +but the raw data itself does not exist as a BIDS dataset, +then the minimally processed data MAY be marked as a raw BIDS dataset. +When represented as a raw BIDS dataset, +the [`desc` entity](../appendices/entities.md#desc) is not needed to distinguish the data and MUST NOT be used. + +The description of the processing in the derivative MUST be clarified in the [descriptions.tsv](common-data-types.md#descriptions-tsv) file, +with at least two columns for desc_id and the actual description. +Other columns MAY be added but are at this moment not standardized. +The [descriptions.tsv](common-data-types.md#descriptions-tsv) file SHOULD include sufficient information to document the details of the data processing that resulted in the derivative. + +The `_eeg.json`, `_meg.json`, or `_ieeg.json` sidecar files MUST be replicated alongside the respective derivative `_eeg.`, `_meg.`, or `_ieeg.` data files, +so that they can be processed as if it were a raw BIDS dataset. +The json sidecar file must be compliant with those for raw data and SHOULD NOT include fields that are specific to the processing that was done on the data. + +## MaxFiltered MEG data + +to be done + +## Other sections (to be done) + +to be done +