WIDE CHARACTERS

[MehdiKarimi]

Is it possible to support wstring in a dicom tag-value?
For example is it possible to set the patient name to a string of wide character type or Unicode type?

[Marco Eichelberg]

As far as the DICOM standard is concerned, there is a very elaborate concept on how non-ASCII character sets are handled through (0008,0005) Specific Character Set and, in some cases, ISO 2022 escape sequences. The character sets typically supported by systems in a specific country varies, for example, most systems in Europe will handled ISO 8859-1 but not anything else. Unicode is supported in UTF-8 representation. If you are talking of the wide character concept used by the Windows OS (i.e. UCS-16), that is not supported in DICOM.

As far as DCMTK is concerned, character strings are just sequences of bytes and their interpretation needs to be handled on application level. The only possible problem concerning multi-byte character sets would be zero bytes within the character string, which should in theory work, but have probably never been tested. Note that such strings never occur in ISO 8859 or UTF-8.