Generating volume datasets

[steini]

Hello,

I want to extract the images of a DCM file directory and generate an array as volume dataset for our own rendering engine. For single files this works via DicomImage.
How do I proceed using a DCM file directory ? What Object and Methods do I need for this ?

Regards,
Marc

[Jörg Riesmeier]

What do you mean by "DCM file directory"? A DICOMDIR file or a filesystem directory with DICOM files? For option A see FAQ #30, for option B see OFStandard::searchDirectoryRecursively().

[steini]

Hello Jörg,

thanks for your answer. I want to connect to a DICOMDIR file, and I did already. Now, I got some problems to access the records. When I understand things right, i have to navigate e.g. with nextSub(). However, when I want to access now the pixel data of a record, e.g.

record->findAndGetUint8Array(DCM_PatientID, pixelData);

then I do not get any results in pixelData, although a print() of the record shows the corresponding pixel data.

Best regards
Marc

[Jörg Riesmeier]

I guess that the pixel data element is contained in a sequence item. In this case you need to set the parameter "searchIntoSub" to OFTrue. That means, the code line would look like this:

Code: Select all

record->findAndGetUint8Array(DCM_PixelData, pixelData, NULL /*count*/, OFTrue /*searchIntoSub*/);

[steini]

Hi Jörg,

ok. Thanks, that works. Now, (I hope) my last problem is, that I got CT as well as PET data in the entire dataset. How can I distinguish between them to visualize just one of them ?

Best regards
Marc

[Marco Eichelberg]

The SOPClassUID and Modality attributes should allow to reliably separate CT and PET datasets.