View Issue Details
|ID||Project||Category||View Status||Date Submitted||Last Update|
|0003255||Slicer4||Module DICOM||public||2013-07-26 11:47||2018-05-29 23:18|
|Target Version||backlog||Fixed in Version|
|Summary||0003255: Create DICOM Seris output has multiple standard non-compliance issues|
As reported by David Clunie's dciodvfy:
Error - Value invalid for this VR - (0x0008,0x0030) TM Study Time TM - Trailing character invalid for this VR = '' (0)
Here's the dcmdump output (some errors are obvious there):
Used TransferSyntax: Little Endian Explicit
(0002,0000) UL 258 # 4, 1 FileMetaInformationGroupLength
Used TransferSyntax: Little Endian Implicit
(0008,0008) CS [ORIGINAL\PRIMARY\AXIAL] # 22, 3 ImageType
|Tags||No tags attached.|
Hi Jim -
You inherited the Create A DICOM Series from Ol' Bill so I assigned this to you. Have there been any fixes at the ITK/GDCM level that might address this?
Longer term as part of the U24 (assuming it's funded) we'll need to make another pass at all the DICOM export functionality in order to smarten things up.
For example, David's dciodvfy code probably wasn't able to detect that the image data is most likely an MR that was exported with a Modality of CT ;)
Steve, Jim - part of it is actually a lot simpler than wrong modality: something in the ITK guts adds extra trailing character after the tag value (for example, look at AccessionNumber tag above).
Hi Andrey - I was actually arguing that it's actually a lot more complicated than simply the extra character or the modality. But fixing the trailing character is a simple fix.
Do we know that this causes any issue in practice? Do PACS refuse to accept the images with the trailing spaces?
Steve: I agree there's a lot more.
I sent the resulting DICOMs to Robert Cormack, he was going to let me know if their system accepts those.
Can't speak to current DICOM libraries, but at one point in time CreateADICOMSeries was "designed" to have the smallest set of DICOM tags needed to import into several PACS and the AW.
Note that DICOM validator tools are great to identify problems wrt the standard. But over the years our biggest problem was figuring out what needed to be in the DICOM to have it accepted by the PACS and AW. For instance, some systems require patient birthdate to be a valid date or it won't import the series.
I'll take a look at the validator output.
Jim: here the validator complains (in particular) about the characters that cannot be present for certain VR types, look at the StudyTime content as an example - the value is invalid. This particular issue appears to be very basic.
Retargeting all old issues to backlog.
|2013-07-26 11:47||fedorov||New Issue|
|2013-07-26 11:47||fedorov||Status||new => assigned|
|2013-07-26 11:47||fedorov||Assigned To||=> pieper|
|2013-07-29 11:20||pieper||Assigned To||pieper => millerjv|
|2013-07-29 11:23||pieper||Note Added: 0009252|
|2013-07-29 11:23||pieper||Status||assigned => acknowledged|
|2013-07-29 12:01||fedorov||Note Added: 0009254|
|2013-07-29 12:47||pieper||Note Added: 0009259|
|2013-07-29 12:57||fedorov||Note Added: 0009260|
|2013-07-29 14:13||millerjv||Note Added: 0009263|
|2013-07-31 06:33||fedorov||Note Added: 0009297|
|2017-02-16 11:24||fedorov||Relationship added||related to 0004344|
|2018-05-29 23:18||lassoan||Target Version||=> backlog|
|2018-05-29 23:18||lassoan||Note Added: 0015779|