Some experiments do not have ScannerType written by XNAT

[Andy Shieh]

Dear XNAT community,

I've noticed in our XNAT instance (version 1.8.3, build: 5) that some experiment session have the Scanner Type field properly written based on the DICOM headers while some do not.

All of these experiments were uploaded using the session import service API in the same way and all of the DICOM files have scanner model information written in the ManufactureModelName DICOM header.

Has anyone encountered issues like this before & know what might be the problem here?

What exactly is the behavior of XNAT writing the ScannerType information? Does it make sure that all DICOM headers have the same ManufactureModelName before writing it into ScannerType, or does it use the majority observation and ignore the minority of the files that have a different ManufactureModelName?

Thanks,

Andy