Fatal error of DCMODIFY

All other questions regarding DCMTK

Moderator: Moderator Team

Post Reply
Message
Author
qxu
Posts: 2
Joined: Thu, 2005-10-27, 18:14
Location: Mountain View, CA, U.S.A

Fatal error of DCMODIFY

#1 Post by qxu »

Hi,
I am trying to modify some Dicom Header information using Dcmodify.
However, in window XP, dcmodiy encounters critical error. And the a dialog box of windows pop up, saying "windows has encountered critical error.....send report....".
In Linux side, the program just hanged up with no response and I have kill it.

I ve run the dcmodify in debug and verbose mode as well. It seems tat it has problem while writing the "new" data back to the original filename.

And by the way, on and off, the dcmodfiy give give me warning -- "metaheader version not supported".

A part of my Dicom Meta Header info is:

[quote]# Dicom-Meta-Information-Header
# Used TransferSyntax: LittleEndianExplicit
(0002,0001) US 0 # 2, 1 FileMetaInformationVersion
(0002,0002) UI =DigitalMammographyXRayImageStorageForProcessing # 30, 1 MediaStorageSOPClassUID

Dcmodify works perfectly fine with all the other dicom images I have worked on. this image comes from FUJI.

What might be the reason that causes this problem? Is it the configure of dcmodify or the problem lies with the Dicom image?


Thank you very much.

Thomas Wilkens
DCMTK Developer
Posts: 117
Joined: Tue, 2004-11-02, 17:21
Location: Oldenburg, Germany
Contact:

#2 Post by Thomas Wilkens »

Please send the image in question to dicom/at/offis/dot/de and we will try to find out why dcmodify doesn't cooperate. Please also provide us with the exact command line you are using and which makes dcmodify hang/fail.

qxu
Posts: 2
Joined: Thu, 2005-10-27, 18:14
Location: Mountain View, CA, U.S.A

#3 Post by qxu »

Hi
I am sorry that I couldnt pass the image to you due to various reasons.
However, I used visual studio to trace the bug and fixed it temporarily.
I belive there is something wrong in "metaheader version" part, dcfilefo.cxx.
The program detects the header version for the dicom image.
If the header version is not supported, the variable pointer "currVers[]" is null.
So it is a bad pointer and it causes exceptions in later while the program tries to read the value that the pointer actually points to.

What I did is, instead of trying to get the value of this "null pointer". I juz print out the version is not supported and dont really try to display the "unsupported metaheader version".

This solves the problem for now because it does cause any exception.
But I guess the problem with the null pointer should still be fixed.
Just some of my observations.

Thank you very much.

Thomas Wilkens
DCMTK Developer
Posts: 117
Joined: Tue, 2004-11-02, 17:21
Location: Oldenburg, Germany
Contact:

#4 Post by Thomas Wilkens »

Please post the exact source code modifications and we will look into the problem and correct the implementation of DCMTK if necessary.

Post Reply

Who is online

Users browsing this forum: Bing [Bot], Google [Bot] and 1 guest