-
Notifications
You must be signed in to change notification settings - Fork 22
Shanoir Exchange Format (SEF)
The idea behind the Shanoir Exchange Format (SEF) is to simplify and standardise data exchange around Shanoir. This means to standardise imports into Shanoir and exports from Shanoir, and exchanges of data between Shanoir instances: Shanoir-old and Shanoir-NG.
As in the future we want to support BIDS exports and imports in Shanoir-NG, the basic idea is to inspire the SEF from BIDS and extend what is necessary for Shanoir exchange. The handling of this format should be put into a separate library to be used by ShanoirUploader and Shanoir servers in the future.
This page details the format.
In a general manner, we could add a "CODE"/hash attribute for every entity, different from ID but still unique for every entity (at least inside a study) to define them. (could be the name ?) As we do this, the user doesn't have to know the ID of the subject/center/study/[...] which can differ and can be generated automatically.
BIDS: participants.tsv (will list all patients and their relevant data, hash values etc.)
participant_id common_name sex birth_date age languageHemisphericDominance manualHemisphericDominance imagedObjectCategory personal_comments hashs...
sub-0d2047861c37b5883cf6cb546ff2436808fbd708b63cb45b9c8b49ae62c7230b M 25 23423...
(sub-xxx contains the subjectIdentifier == hash generated with pseudonymus or algo Neurinfo)
All these infos go into that file coming from ShUp:
- patientBirthDate 01/01/2011
- firstNameHash1 345...
- firstNameHash2 345...
- firstNameHash3 345...
- lastNameHash1 345...
- lastNameHash2 345...
- lastNameHash3 345...
- birthNameHash1 345...
- birthNameHash2 345...
- birthNameHash3 345...
- birthDateHash 345...
Quite some informations are missing in BIDS format to be able to export/import an examination. We can add a "examination_description.json" file with following informations: EDIT: could be scans file enriched with specific field https://bids-specification.readthedocs.io/en/latest/03-modality-agnostic-files.html#scans-file
- Center (CODE)
- Examination_Date
- Comment
- Subject_weight
- Link_to_other_files[]
- (+ potentially pre-clinical fields to be defined)
We also need to add information about which study we are working on. Use dataset_description.json We can add a mandatory field with the CODE/HASH/NAME of the study we are working on.
?
All datas are more or less in the data we are importing/exporting. One question remains: How to split one file (DICOM) in multiple datasets (in case of multiple datasets for one examination) => Export and import Also, if we have a DICOM file in entry, we have to convert it to nifti when exporting.