From 777064cf63c71cde1d5208273956221898855009 Mon Sep 17 00:00:00 2001 From: Stefan Appelhoff Date: Tue, 7 Nov 2023 11:07:34 +0100 Subject: [PATCH] remove part of the language --- src/derivatives/common-data-types.md | 10 +--------- 1 file changed, 1 insertion(+), 9 deletions(-) diff --git a/src/derivatives/common-data-types.md b/src/derivatives/common-data-types.md index 908d9d2138..b1ac304fc8 100644 --- a/src/derivatives/common-data-types.md +++ b/src/derivatives/common-data-types.md @@ -253,18 +253,10 @@ All REQUIRED metadata fields coming from a derivative file’s source file(s) MU be propagated to the JSON description of the derivative unless the processing makes them invalid (for example, if a source 4D image is averaged to create a single static volume, a `RepetitionTime` property would no longer be relevant). -As each file includes what was computed by increment, -a trail of the computational steps performed is created. -Because many steps can occur until the fully clean data are obtained, -and because it is OPTIONAL to save every file or step, -the simpler `desc-preproc` and `desc-proc` are available. -It becomes however important to document what such pre-processing or processing are, -and record the order of computational steps. -This can be done in the JSON sidecar files or alternatively described in a `descriptions.tsv` file. ## descriptions.tsv -To keep a record of processing steps applied to the data, a `descriptions.tsv` file can be used. +To keep a record of processing steps applied to the data, a `descriptions.tsv` file MAY be used. The `descriptions.tsv` file MUST contain at least the following two columns: - `desc_id`