Posts by msa

Meanwhile, I have found another type of header corruption: Some images contain a complete metaheader. The transfersyntax says that the file is encoded in LittleEndian (don't remember if it was implicit or explicit, but I could try to find out if it helps). However, the pixeldata is obviously compressed. I find a hint in the DerivationDescription about the used compression algorithm.

Following Mr. van Herk's suggestion, I would like to transfer the images to our server using DICOM-C-Store. But is there any way of making sure that all images were successfully transferred? Does Conquest support storage commitment as an SCU? Is there a mode which tells conquest to delete the images after successful transmission? Or are there any other ideas how to accomplish the transfer?

Thanks in advance

Markus

Hi all,

I have had a small discussion with Mr. van Herk about this DICOM files that reside in the Conquest filesystem:

Quote from msa

Dear Mr. van Herk,

one of our customers asked us to import images from a conquest server. Our application is based on the OFFIS DICOM Toolkit (DCMTK) and we want to read the images recursively from the filesystem to send them to another server afterwards.

It works fine with some of the images but there are also some of them which seem to have an imcomplete metaheader. The metaheaders of the affected images I examined only contain the transfersyntax-UID and nothing else. The customer says that these images have been received by the conquest server after the compression was de-activated, but I am not sure if he is right.

So my questions are:
1) Are my observations plausible to you?
2) If so, what would be your suggested approach to export all images from the conquest-server, given that it should be possible to determine which images were successfully exported and which failed?

Display More

Quote from Marcel van Herk

If the images are stored as .v2 files then the metaheaders are stripped in the latest version. Files with .dcm extension keep the metaheader. I never noticed that the length is missing, but it can be true since the length is also created automatically by the UCMDC libraries when reading the images. Could you also send me some sample images that work and fail? We use dcmtk 3.5.4 to compress and decompress jpeg data and that works, but then we store the data as dcm not v2.

To solve the problem you could maybe transmit the images from conquest into your server.