[Dcmlib] Fwd: Series Not Recognized In Some Software

[Mathieu Malaterre]

FYI. GDCM 2.x does not support those image, I'll try to find a pseudo
fix. I did not check GDCM 1.x

---------- Forwarded message ----------
From: Mathieu Malaterre <mathieu.malate... at gmail.com>
Date: Sep 28, 8:10 pm
Subject: Series Not Recognized In Some Software
To: comp.protocols.dicom


On Sep 27, 9:49 pm, lintrad... at gmail.com wrote:

> Hello,

> I am having trouble with some software (particularly eFilm and
> ClearCanvas) recognizing that a set of files belongs to the same
> series and should load them together - they are viewed as a group of
> individual files. I am puzzled as nothing immediately jumps out to me
> as incorrect,

You are kidding right ? Those are among the worst DICOM file aver, the
bug are extremely painful to deal with:

Let's start with the easy part, incorrect VR:

(0008,0030) CS (TM) [100000.000]                         # 10,1 Study
Time
(0008,0032) CS (TM) [100000.000]                         # 10,1
Acquisition Time
(0008,0070) CS (LO) [TOSHIBA ]                           # 8,1
Manufacturer
(0008,0090) CS (PN) [???]                                # 4,1
Referring Physician's Name
(0008,1030) CS (LO) [Anonymized with DicomWorks]         # 26,1 Study
Description
(0008,1090) CS (LO) [Xvision ]                           # 8,1
Manufacturer's Model Name
(0018,1130) CS (DS) [+90.00]                             # 6,1 Table
Height
(0020,0037) CS (DS) [-1\0\0\0\-1\0]                      # 14,6 Image
Orientation (Patient)

And best of all, the Pixel Spacing not specified properly:

(0028,0030) DS [0.468 0.468 ]                             # 12,2 (1)
Pixel Spacing

I can bet you almost anything that only the first value is read, while
the second can never *ever* be accessed.

>  and that some software does recognize them OK.

Please gives us this list of buggy software. They are absolutely not
compliant. The only thing I can think of is that they default to
isotropic image, and use the first value of Pixel Spacing to find the
second one.

-Mathieu