Open neurolabusc opened 4 years ago
@neurolabusc if there are acquisitions varying with respect to TR
, the convention is to put each value in separate jsons per image. In this case, Inversion Time
looks OK as is.
But if Inversion Time
and Inversion Times
are mutually exclusive, then we may note that.
Agreed. Prior versions of dcm2niix followed BIDS and only populated InversionTime
if 0018,0082 exists, the upcoming release will populate it if either 0018,0082 exists or if 0018,9079 exists with a single value. This should help users of Siemens XA systems. I do not have access to Siemens XA, so I am not sure if it ever creates an enhanced DICOM where more than one TI is listed for a single file. Since I have no exemplars, dcm2niix will generate a warning if this ever occurs.
My only comment here is that I think the bmp should be updated to list multiple possible sources for this useful parameter.
Thanks @neurolabusc, I agree! I will make a note for this case.
The specification suggests InversionTime should be sourced from DICOM tag 0018,0082. However, some modern enhanced DICOMs store inversion time in tag 0018,9079. The latter allows an array of multiple values, but some vendors (e.g. Siemens XA) use this for single values as well. I would update the document to list either source.