DICOM @ OFFIS

last question is -Wall for cxx flags in cmake configuration

I've already forwarded this question to the DCMTK team members who prefer to use CMake (on non-Windows systems).

So, the issue (crash) is caused by an incomplete modification of DCMTK's source code, right? anyway, i think it is good idea to add "default" Yes, one could do this, e.g. you could/should do this as part of your fork, but I am not sure whether this really makes sense for the public DCMTK source code...

it crashed in dcmdata/libsrc/dcpixel.cc:1195 Do you really use "master"? http://git.dcmtk.org/?p=dcmtk.git;a=blob;f=dcmdata/libsrc/dcpixel.cc;h=7d1e61550cb2b85db3d72d58955b65063e2481dc;hb=HEAD#l1195 1195 return (*current)->pixSeq->loadAllDataIntoMemory(); It's line 1168 in "master". Anyway, I teste...

I understand what YBR_RCT​ and YBR_ICT are only allowed for compressed images but it seems not all developers understand it also, so, our equipment sometime receives such dicoms from some "client tele-medical environment" Here's the reference to the relevant section in the DICOM standard: http://di...

Which version of the DCMTK did you use? As far as I can see, there is no such thing as EPI_YBR_RCT or EPI_YBR_ICT (enums) since these photometric interpretations (YBR_RCT​ and YBR_ICT) are only allowed for compressed images, not for uncompressed images. The same is true for YBR_PARTIAL_420. switch (...

Does this commit (from May 2019) solve your issue?

The DICOM standard says with regard to the SOP Instance UID: same value means same DICOM object, different value means different DICOM object. In your case, the filename (Referenced File ID) is not relevant but the value of SOP Instance UID.

Short question: did you change the SOP Instance UID of the DICOM file before the second addition?

Below is my code Snippet.

Where?

Hmm, I tried the same calls with DCMTK 3.6.6 (current release), and it works. > storescp --version $dcmtk: storescp v3.6.6 2021-01-14 $ storescp: DICOM storage (C-STORE) SCP Host type: x86_64-unknown-linux-gnu Character encoding: system default (unknown) External libraries used: - ZLIB, Version 1.2....

By default, the call of loadFile() does not load all element values into main memory, only those that are smaller than 4 kbytes.
You could either increase this value (parameter "maxReadLength" of the method) or call loadAllDataIntoMemory() before saveFile().

I am talking about thousands of studies. Is there a way to dump the output of findscu to file which can be passed to movescu as a [dcmfile-in...] parameter? findscu's --extract option should do what you want. However, you have to remove all non-Unique Keys from the C-FIND Response datasets before f...

You cannot filter by Manufacturer using C-MOVE (by the way, also PatientName should not work). Only the Unique Keys of the respective levels (PATIENT, STUDY, SERIES, IMAGE) are allowed, i.e. PatientID, StudyInstanceUID, SeriesInstanceUID and SOPInstanceUID. If you want to filter by Manufacturer, you...

The Overlay Plane Module is not defined for the Segmentation IOD (in fact, it is forbidden for this IOD), so what are the "overlays" you are referring to?

I tested it on my Ubuntu 20.04 (LTS) machine with latest updates and got the following (on SCU side): echoscu-tls: dl-call-libc-early-init.c:37: _dl_call_libc_early_init: Assertion `sym != NULL' failed. Aborted (core dumped) Here is the --version output: $dcmtk: echoscu v3.6.6 2021-01-14 $ echoscu: ...

Since the Visible Light Image IOD is a single frame IOD, i have to write each captured image to a file and send it the server. In fact, there are also VL Image IODs for video data but since your images are no videos it make no sense to use one of them (as far as I understood your use case). Do i ha...