Mapping subjects with overlays

[suliciac]

I've detected an issue with the subjects mapping when the run has more than one localizer of the same subject.

If we want the data to by anonymized, the Patient ID cannot be used as BIDS subject (in this case, the mapping works as it should be). However, when selecting a numerical ID, it fails to associate all sequences in the same subject, creating multiple subjects.

@bendhouseart would it be feasible to address this?

[bendhouseart]

I would say to exclude localizers as we don't include those with BIDS, but that doesn't explain your issue here. Would you be able to show me how your unconverted data is organized? Or walk me through a conversion?

[suliciac]

Well, the disturbing files are not the presence of multiple localizers but reconstruction images or overlays in general, I'd say.

Maybe they could be identified by some optional attributes in the DICOM file only present in this type of image:

• (60xx,xxxx) Overlay Plane
• (0008,2111) Derivation Description