Skip to content

bidsmri2nidm: capture BIDS field maps with a FieldMap image-usage type - #452

Merged
dbkeator merged 1 commit into
incf-nidash:masterfrom
dbkeator:master
Aug 13, 2026
Merged

bidsmri2nidm: capture BIDS field maps with a FieldMap image-usage type#452
dbkeator merged 1 commit into
incf-nidash:masterfrom
dbkeator:master

Conversation

@dbkeator

Copy link
Copy Markdown
Contributor

Captures BIDS field-map acquisitions that were previously dropped or mislabeled, validated to produce graphs isomorphic to the LinkML refactor via typed-shape A/B on real ABIDE data.

Build the pybids layout with validate=False so real-world non-conformant files (e.g. ABIDE II field maps placed under dwi/) are captured rather than silently dropped by BIDS validation.
New nidm:FieldMap image-usage type (Constants.NIDM_MRI_FIELDMAP) plus BIDS_Constants entries for the fmap datatype and field-map suffixes (fieldmap, magnitude, magnitude1/2, phasediff, phase1/2, epi). Field-map scans are now labeled FieldMap instead of mislabeled DiffusionWeighted. (Vocabulary term proposed to nidm-experiment separately.)
Emit the BIDS run index in the anat branch (it was only emitted for func/dwi).

Validation: typed-shape parity (scripts/parity_compare.py) legacy vs LinkML — ABIDE Caltech PARITY OK (6,635 instances) and ABIDE II NYU_1 PARITY OK (42,211 instances; exercises DWI, field maps, multi-run, multi-session).

@dbkeator
dbkeator merged commit 73c32b2 into incf-nidash:master Aug 13, 2026
6 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant