Request for DICOM images from China

1

I am trying to improve dcm2niix DICOM to BIDS conversion for international datasets. By default, DICOM saves text using ISO_IR 100, but the BIDS sidecars use JSON that requires ISO_IR 192 (aka UTF-8). While DICOM images can specify different datasets, dcm2niix currently assumes that the DICOMs use ISO_IR 100. To date, all the DICOM data I have seen from GE, Siemens and Philips uses ISO_IR 100, and while the UIH data I have seen uses ISO_IR 192, this dataset is restricted to Latin characters (so I can not validate conversion of non-Latin text).

My question to international DICOM users (China, Korea, etc) is:

  1. Are details like patient name and institution name always stored in your DICOM as Latin (alphabetic characters), or do you sometimes use ideographic characters?
  2. If the latter, could you share sample DICOM images with me (the images could be of a phantom, not a human)?
1 Like